Session Chair Reports

Impact Assessment in the Age of Artificial Intelligence

This information will inform the closing plenary, proceedings, and future planning.

SESSION NAME

SIA as the basis for ongoing participatory monitoring: examples from around the world

NAME(S) OF SESSION CHAIR(S):

Kathy Witt, Will Rifkin

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI could be useful for ongoing monitoring data collection, e.g. social media monitoring

1B. WHAT SPECIFIC EXAMPLES FROM YOUR SESSION HIGHLIGHTED THE ROLE OF AI IN IMPACT ASSESSMENT, PARTICULARLY THE OPPORTUNITIES IT OFFERS AND THE CHALLENGES IT PRESENTS IN SHAPING POLICIES, STRATEGIES, AND PROJECTS?

not really discussed

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Social impacts occur and can change over time. It is important to continue monitoring impacts after the project has been approved and goes ahead. Measure things that are important to local people.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Think about including indicators in SIA that can be tracked over time. Consider using participatory methods to choose the indicators so they reflect what is important in that region/community.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

While agreed that ongoing monitoring is important, it can be difficult to secure funding that will support monitoring activities over time. Consider establishing formal structures that support ongoing monitoring within the project approvals and planning processes.

SESSION NAME

Advancing Biodiversity Assessment, Risk Management, and Conservation Strategies

NAME(S) OF SESSION CHAIR(S):

Mark Leckie

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

That transparency is absolutely paramount of metadata and methodology. AI should be approached with caution for modelling especially.

No AI users.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

There was alignment across speakers in methodology and transparency, and this data was making real on the ground decisons.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Use high quality data and rigorous transparent methodologies

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

N/A

SESSION NAME

Meaningful Stakeholder Engagement: For and With Affected Stakeholders

NAME(S) OF SESSION CHAIR(S):

Karin Buhmann and Alberto Fonseca

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The topic of IA was not directly addressed by any presenter. However, one of the presenters of the 3rd slot raised a good point, i.e., that AI is just a tool that can enhance or weaken public participation.

The fact that AI was almost fully absent of the presentations and even of the discussions is itself an evidence that the session topic (meaningful stakeholder engagement) is still not directly influenced by AI. While AI, of course, can affect public participation, it does not seem to be yet a relevant issue for the researchers and practitioners who attened the session.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The session had numerous relevant messages and lessons from many case studies and policies mentioned in the presentations. Overall, the lessons reveberate the takeaways of the Routledge Handbook on Meaningful Stakeholder Engagement, co-edited by the session chairs.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

As mentioned in the session, the Routledge Handbook on Meaningful Stakeholder Engagement does synthesize the state of the art of MSE.. The book is availabble here: https://www.routledge.com/The-Routledge-Handbook-on-Meaningful-Stakeholder-Engagement/Buhmann-Fonseca-Andrews-Amatulli/p/book/9781032482675?srsltid=AfmBOoopH42ozYjpOyln7Er8qIyJiqkjC1D8zdMogdK1wOMqul3Kjl7F

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Please see 3 above.

SESSION NAME

Meaningful Stakeholder Engagement: For and With Affected Stakeholders

NAME(S) OF SESSION CHAIR(S):

Karin Buhmann and Alberto Fonseca

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Please see 3 above.

SESSION NAME

Meaningful Stakeholder Engagement: For and With Affected Stakeholders

NAME(S) OF SESSION CHAIR(S):

Karin Buhmann and Alberto Fonseca

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The fact that AI was almost fully absent of the presentations and even of the discussions is itself an evidence that the session topic (meaningful stakeholder engagement) is still not directly influenced by AI. Whille AI, of course, can affecte public participation, it does not seem to be yet a relevant issue for the researchers and practitioners who attened the session.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Please see 3 above.

SESSION NAME

Mainstreaming AI in environmental Baselining and Monitoring

NAME(S) OF SESSION CHAIR(S):

Pippa Howard, Noelia Jimenez Martinez

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

We noticed a big uncertainty on how AI will transform the way we asses baselining and monitoring for impact assessment. Mainly the concerns on how the models are black boxes for most practitioners, and currently 'autonomous' AI systems (like ChatGPT), offer little to no accountability. The panel considered crucial to have 'humans in the loop', and the correct data representing ’reality’ -- from granular species level data to indigenous know-how, and global affairs’ information, included in the training of AI systems. At the same time people were optimistic that if engaging effectively with these new technologies, many disruptions and optimizations were possible, leaving a positive net effect in how we manage impacts and audit complex systems, at scale and in real time. All the benefits of AI and nature tech can help provide the rigour, science, evidence, integrity, accountability and audibility needed to underpin and enable positive outcomes from this somewhat-vulnerable process.

All presenter ins the session provided examples of new technologies used to map biodiversity and changes in the landscape. William Paterson presented his work with drones for mapping species in the Red Sea. Marissa Murphy presented the case environmental assessment for offshore baselining and how different tooling could make the process more efficient. Noelia Jimenez Martinez presented the case of eDNA and geospatial data integrated to understand where most irreplaceable species can be found, to and be decide the best conservation and restoration areas in mining and oil industries. The discussion with the panel and participants covered the use of integrated datasets, including all these techniques, plus others, such as camera traps, bioacoustics, foundation models, LLMs, etc, to complete the assessments and track changes in real time. The challenges highlighted were: access to the ‘right’ data, open source data, interoperability of the data, training and understanding how these new tools work, integrating all parties, representing the communities, the insights offered by these tools, how to deal with complexity, and the ultimate decision making processes based and derived from these practices.As new technologies emerge, it's crucial that we build capacity among practitioners to understand both the appropriate uses and potential misuses of these tools. There's always a risk that AI could be co-opted by bad actors, particularly in ways that support greenwashing. To maintain the integrity of impact assessment, we need to stay vigilant—ensuring transparency, accountability, and critical evaluation of how these technologies are applied.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

As new technologies emerge, it's crucial that we build capacity among practitioners to understand both the appropriate uses and potential misuses of these tools. There's always a risk that AI could be co-opted by bad actors, particularly in ways that support greenwashing. To maintain the integrity of impact assessment, we need to stay vigilant—ensuring transparency, accountability, and critical evaluation of how these technologies are applied.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

To stay updated and keep a growth mindset, to find and stay connected with trusted experts and diverse voices who are actively engaging with emerging technologies—especially those who understand both the technical and ethical dimensions of AI. It’s also important to build your own literacy in these tools—not necessarily to become a coder, but to be a critical and informed user. Encourage interdisciplinary collaboration, and don’t shy away from questioning the assumptions baked into the data and models you use. Finally, remain grounded in the core values of impact assessment: transparency, climate justice, inclusion, and sustainability. Technology should support these principles.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

For policymakers, users and producers of AI tools, the key recommendation is to develop regulatory frameworks that are agile enough to keep pace with AI innovation, while still upholding transparency, accountability, and environmental justice. This includes requiring disclosure of methodologies and data sources used in AI-driven assessments, and ensuring that AI tools do not replace—but rather support—public consultation and expert review. For community and Indigenous stakeholders, mechanisms must be in place to support meaningful participation, including the right to challenge or interpret outputs generated by AI models. Their knowledge should be considered not just supplementary, but central to the design and evaluation of any AI-driven impact assessment process. Across all stakeholder groups, there is a shared need to build capacity, promote data equity, and actively address biases in both datasets and model design.

SESSION NAME

Bridging the Digitl Divide: enhancing EIA in low and middle-income countries

NAME(S) OF SESSION CHAIR(S):

Chair - Landry Fanou , Co-Chair - Anna Chadarevian

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The need for contextually and culturally aware use of digitalisation in AI in low middle income countries. This call in most case on hybrid approach for the deployment of digital infrastructure and also for the need for IA for the deployment of IA.

disruption of labor market in garment industry in south east Asia and social media in Kenya and stakeholder engagement in Nigeria.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The enabling environment and capacity of stakeholder and the EIA process holistically should be integrated while thinking of AI system in LMIC.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Be culturally and contextually aware of the use of digital tools and AI IN EIA processes . Be mindful abt potential risk of exclusion of stakeholders given the digital divide.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Policymakers need to integrate and update their policies to have proper use of AI. Foresee the unemployment implication and design policies to address these social aspects.

SESSION NAME

Use of data, technology, and AI tools for enhanced accountability

NAME(S) OF SESSION CHAIR(S):

Benoit Vanwelde and Laurence Levaque

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Presenters highlighted that AI should be considered as a complementary tool, and not as a solution.

Participants agreed that AI can enhance accountability by streamlining complex processes, improving data review and organization and enabling more efficient monitoring. Discussions however underscored ethical concerns about training data used for AI applications, and the need to keep human oversight in preventing misinformation and misuse.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Participants to the discussions organised around the three round tables explored strategies to extend AI’s benefits to developing regions, ensuring inclusivity rather than reinforcing biases.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Institutional learning also emerged as a hurdle, requiring internal alignment within our respective institutions for effective use and implementation. Investments and trainings are needed and should percolate down to each users of AI tools.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Cases of unequal power dynamics between stakeholders, including communities that are based on oral cultures or with limited digital access were discussed.

SESSION NAME

Assessing the health impacts of AI

NAME(S) OF SESSION CHAIR(S):

Margaret Douglas, Liz Green

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Use of IA could have both positive and negative impacts on health, for example: Use of IA in the healthcare sector could improve healthcare outcomes but benefits may not be equitably realised and there is also potential for overdiagnosis Use of AI in the manufacturing sector could reduce costs, reduce environmental impacts, reduce wastage but could also lead to unemployment with wide ranging negative health impacts for affected communities In all sectors energy use for AI will have environmental and health impacts There are also data privacy concerns in AI use across sectors

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The importance of context in considering impacts The importance of considering distributional impacts

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

SESSION NAME

Assessing the impacts of food systems: experiences from the UN RBAs

NAME(S) OF SESSION CHAIR(S):

Arundhati Inamdar Willetts, Maria Rita Borba

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The participants discussed and explored key aspects that should have been taken into consideration when the project used in the case study was implemented in the 1970s, and that, thanks to change in regulation and international practice have been incorporated in recent projects. Participants also discussed that, despite the challenges and flaws in current projects, it's recognizable that advancements have been made in terms of policy, application and enforcement of better ES practices.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

For practitioners that are working particularly in the development/humanitarian field, the basic pronciples and aspects of impact assessment should be considered, with special attention to the vulnerability of the communities and their livelihoods. Some of the key points for consideration mentioned during the session included: robust baseline, defined methodology, capacity building (particularly if changing livelihoods activities), involvement of youth, human based approach (right to food has suppressed other rights like land tenure), legacy issues (promises not kept), hearing vulnerable people in the communities, ability to adapt the project design during implementation.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Developers and implementers to get to know and include local culture and decision making practices, for example, working with village chiefs. Best international practices of inclusivity are key but need to be adapted to local contexts and existing strictures to avoid conflicts and clashes, enhancing positive outcomes.

SESSION NAME

Korean Endeavor to Accomplish Digital EIA: Demonstration

NAME(S) OF SESSION CHAIR(S):

Sangyun Lee

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

In order to utilize artificial intelligence in environmental impact assessments, it is a prerequisite to establish an information system. However, in many developing countries, the necessary conditions for this are often not in place. Therefore, when discussing the use of artificial intelligence in environmental impact assessments from an international perspective, support for developing countries should be taken into consideration.

- Contributing to minimizing deficiencies related to environmental impact assessments. - Since complete surveys are not cost-effective, it contributes to the selection of survey points

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

- "When artificial intelligence does not always tell the truth, it raises concerns about trust in the environmental impact assessment system

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

- When building a digital EIA (Environmental Impact Assessment) system, it is important to develop it based on field experience; therefore, the participation of practitioners is essential

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

- Rather than developing vague digital EIA-related policies, it is necessary to take a more pragmatic approach that carefully considers the specific conditions of each country

SESSION NAME

Essential AI Learnings: ESIA Tools from Multilateral Financial Institutions

NAME(S) OF SESSION CHAIR(S):

Hak Joo Song, Brian Rafor

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Indeed AI is a big help in the workflow of the impact assessment. There is a wide application of it from the typical use case of reviewing and summarizing relevant impact assessment documents up to sophisticated ones like fusing in geospatial application. The integration of AI streamlines processes, enables rapid access to relevant information, facilitates the exploration of diverse insights, reduces human error, and enhances the accuracy of predictions and models—ultimately supporting more informed decision-making.

Each of the organization’s presentation provided numerous examples of how AI has been utilized in their operations. Key points included the use of AI-assisted Geospatial tools for E&S Due diligence and Smart Monitoring by IADB, Generative AI and Multi Source Data Streams for Risk Identification and Baseline Information by FAO, the combination of Geospatial Tools and E&S Agents for Risk Screening & Classification by ADB, and the C-ESRS AI Assistant for the zero-draft generator of the World Bank and the Geospatial ESF Risk Assessment toolkit for the E&S risk screening of the World Bank. These tools illustrated how AI can help integrate operational processes and technology for efficiency improvements and address resource limitations, by bringing external and internal datasets. The session also discussed the challenges and potential future developments of these tools.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

AI, like any technological change, requires careful implementation assessment. Organizations and individuals may be thrilled by technology's selling points, but it's essential to examine the reasons for its rollout. For example, the session highlighted the importance of maintaining human connections in impact assessments, emphasizing their significance.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

It is recommended to experiment with technologies to assist in our work. The advancement of the field is important. As we conduct the session today, various tools are already available. It is advisable for practitioners to utilize existing tools for their specific use cases rather than starting from scratch unless absolutely necessary.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

A major topic of the discussion was harmonization and data sharing among organizations. The focus is now on improving the performance of these tools by utilizing extensive data and knowledge from peer organizations, including Multilateral Financial Institutions and UN Organizations.

SESSION NAME

Impact Assessment and Governance for Climate Adaptation and Mitigation in Megacities

NAME(S) OF SESSION CHAIR(S):

Yuan Xu, Daniel Guttman

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

N/A

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Impact assessment plays a significant role in the formation of a sustainable climate governance strategy that considers both mitigation and adaptation.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

More quantitative, transparent, objective and verifiable impact assessment can add values.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Climate policies should consider the impacts on various stakeholders to enhance their own sustainability.

SESSION NAME

Session 428 A rights-based approach to IA: Indigenous resilience and resurgence through engaging in IA

NAME(S) OF SESSION CHAIR(S):

Callie Davis-Flett, Mandy Olsgard

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Resources need to be considered - Budget, Time, Skills Processes need to be developed - FPIC, OCAP, Co-development If done incorrectly AI could perpetuate and expedite colonization of Indigenous communities Faster and more efficient isn’t always better

Greatest risk identified was that existing AI platforms perpetuate and embed colonial epistemologies with systematic operationalization of bias against non-white, non-male, and non-Western peoples Solutions identified were Indigenous built AI. Integrating Traditional Knowledge into AI tools is not recommended or possible yet. Need to understand and develop mechanisms to protect TK and communities data.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

AI could be useful to summarize western science IA, EIA, monitoring data that are submitted by industry and government to support Indigenous communities manage workloads.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

AI could be useful to summarize western science IA, EIA, monitoring data that are submitted by industry and government to support Indigenous communities manage workloads. Existing AI platforms perpetuate and embed colonial epistemologies with systematic operationalization of bias against non-white, non-male, and non-Western peoples. If you want to rely on AI, the tools should be co-developed with Indigenous communities if they consent.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Any not-Indigenous stakeholder should not be using AI to assess impacts to communities or putting community data into any existing AI tools. Resources need to be considered - Budget, Time, Skills Processes need to be developed - FPIC, OCAP, Co-development If done incorrectly AI could perpetuate and expedite colonization of Indigenous communities Faster and more efficient isn’t always better If relied on in current state, AI could perpetuate and expedite colonization of Indigenous communities

SESSION NAME

Improving IA efficiency for climate change: why, when and how

NAME(S) OF SESSION CHAIR(S):

Anna Johnston

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

That efficiency cannot come at the expense of efficacy and fairness (the three are interrelated), that governments around the world are under increased pressure to cut corners in the name of economic development and mitigating climate change, and that comparing projects' GHG emissions to climate plans and carbon budgets can be efficient ways to understand climate effects.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

That efficacy can be achieved without sacrificing efficiency and fairness with greater use of regional and strategic assessment, co-governance with Indigenous peoples, and multi-jurisdictional cooperation to the highest standards. To compare projects' GHG emissions against any relevant climate plans and policies, including targets and budgets where they exist.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Not to give up on the promise of impact assessment just because of real or perceived urgency. View efficiency as an efficient way of achieving substantive objectives (e.g., sustainability, a climate-safe future, halting and reversing nature loss) rather than faster assessments. Use pilot projects to test ways to better assess GHG emissions and climate impacts. Continue to evaluate the efficacy of climate assessments in the spirit of continuous improvement.

SESSION NAME

Practical assessment and mitigation approaches in IA

NAME(S) OF SESSION CHAIR(S):

Meredith Henderson

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Nil

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Collaboration and bold policy changes can assist in improving mitigation measures.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Engage with specialists and consider cross disciplinary issues with a desire to find resolutions.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Clear guidelines on mitigation is required from regulators and even understanding the baseline areas required to assess impacts.

SESSION NAME

Bioregional Impact Assessment

NAME(S) OF SESSION CHAIR(S):

Rufus Howard

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

People are excited to explore new concepts like bioregional philosophy in relation to IA.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Look more widely outside of the traditional IA field at related fields such as deep ecology and bioregionalism for new inspiration for innovation in IA.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Many of the concepts and improvements espoused by IA practitioners like better stakeholder participation, better holistic planning, more strategic approaches etc lack mechanisms for implementation. A bioregional planning approach provides a vehicle for many of these approaches to be brought together and applied.

SESSION NAME

Self-determined and culturally relevant Gender Based Analysis Plus and Intersectional Impact Assessment: An invitation for Action

NAME(S) OF SESSION CHAIR(S):

Dawn Hoogeveen, Patricia Fitzpatrick

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Questions about equity and data sovereignty were raised.

N/A

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

-The importance of creating safe spaces for dialogue to share experiences and intersectional practice -Practitioners should continue to integrate intersectional analysis in IA - including analysis outside of silos -Collective action and every day resistance is essential to empower knowledge sharing at all scales -Those with privilege and power need to advocate for those who are more precariously placed

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

-It is important to include requirements for intersectional analysis in IA - as it ensures that it occurs

SESSION NAME

Indigenous Knowledge as a Driver of Innovation in Impact Assessment

NAME(S) OF SESSION CHAIR(S):

Nina Barton, Hereward Longley

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI presents opportunities to support capacity challenges faced by Indigenous communities by potentially increasing the efficiency of data analysis for Indigenous-led assessment. However, it cannot replace decision making and human judgement.

AI presents opportunities for increasing the efficiency of using Indigenous Knowledge databases to identify project interactions and potential effects. AI presents data sovereignty challenges for Indigenous Knowledge (IK). IK is often confidential and contextual and cannot be entered into LLMs.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Indigenous nations are rapidly building impact assessment frameworks and capacity to conduct Indigenous-led assessments. Regulators and proponents should work actively with Indigenous nations to support the development of these processes. Doing so may help improve the relevance of impact assessments to Indigenous communities support FPIC and increase regulatory clarity.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Impact assessment practitioners should approach Indigenous-led assessments with and open and creative perspective. They should work with communities on concordance reviews to determine which components of the impact assessment Indigenous nations want to undertake, and support and work with those communities to undertake their Indigenous-led assessments. Practitioners should look to expand regional assessment tools for off-ramping regional and transboundary issues to allow greater focus on project level assessments.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Any efforts to undertake regulatory streamlining, especially in Canada, should focus on buttressing Indigenous-led assessment processes and not try to bypass Indigenous rights in any efforts to expedite approvals. We suggest that supporting Indigenous-led processes can support consensus building on the accuracy of the assessment, while fast-tracking that ignores Indigenous rights will trigger litigation and conflict that is more time consuming and costly in the long run.

SESSION NAME

What Science and whose Art? Demarcation of AI and Human Roles in IA.

NAME(S) OF SESSION CHAIR(S):

Jack Krohn, Tanya Burdett

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The intersection ("boundary") will continue to move. Human accountability will remain critical for the application of ethics and wisdom even to tasks that might be able to be automated.

No specific examples raised - not the nature of this session.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The output from the session will be a demarcation statement delineating the human and AI roles across different process elements of impact assessment. The statement will be assmebled from session-generated documents and will be shared as a draft through IAIA Hub.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Stay aware of changing AI capabilities relvant to your practice. Maintain healthy scepticism. Accept accountability - don't hide behind "what the computer said".

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Recommendations will be encapsulated in the draft demarcation statement.

SESSION NAME

Impact assessment for mine closure for post-mining transitions

NAME(S) OF SESSION CHAIR(S):

Sara Bjørn Aaen, Izhar Mithal Jiskani

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

N/A

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Key challenges in post-mining transitions: - We need to implement IA throughout the cycle including post-closure, the EIA is too limited to handle the transition issues. And current frameworks is not adequate - Legacy mines present a particular problem as to who is responsible and how to ensure sustainable transition - Mine closure can happen suddenly, long term planning for post-closure is difficult

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

- Due dilligence related to risk of sudden mine-closure is imperative. - IA should consider not only direct, but also indirect and long-term impacts of mine closures

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

- We need new governance frameworks to guide post mining transitions that set requirements for post-closure impact assessment - It needs to embedded in mineral policies on a national and global level

SESSION NAME

Post-resettlement Livelihoods: Reality, Experiences, and Lessons

NAME(S) OF SESSION CHAIR(S):

Gwendolyn Wellmann

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

N/A

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

N/A

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

One has to think of possible impacts far into the future, not just for the foreseeable future and plan for ongoing mitigation.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

In terms of South Africa, government and policy makers should look into relocation of those households who are affected by disasters repeatedly (eg. Flooding, fires) and what these relocation processes should look like.

SESSION NAME

Are Competent Experts and Capacity Building still core to IA in spite of AI?

NAME(S) OF SESSION CHAIR(S):

Josh Fothergill

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

IA professionals and organisations using them will need to place more emphasis on capabilities of staff and how these are developed as AI wraps up knowledge and data related roles, that have up to now been core to training and building human IA experience. Organisations and senior professionals need to set the bar on what is expected and acceptable in using AI in work for more junior staff who are almost certainly more experienced and capable in AI use than their bosses.

Ivar's presentation explained the views of 'traditional' IA professionals on the use of AI in Denmark with two distinct groups emerging Team A (positive adopters) and Team B (concerned individuals who are more resistant). AI training for IA professionals in IFC

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Need to orient to capabilities of IA professionals (behaviours, skills, mindsets) alongside knowledge process, standards, ESH, etc.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

SESSION NAME

Korea's Urban Planning Technology Development in Response to Climate Change

NAME(S) OF SESSION CHAIR(S):

Seunghyun Jung

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Beyond the AI modeling, many researches have done by AI utilization such as research writing, visualization, and presentation. We need to set more rigid criteria for the information judgement.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

More practical way to uitlize and adopt AI in real wolrd.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Research outcome has been provided mainly for government officers, professionals.

SESSION NAME

Can IA be more ‘future-smart’ - incorporating Futures Thinking in IA

NAME(S) OF SESSION CHAIR(S):

Francini van Staden; Nicholas King

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI must not be used in 'blind-faith', it requires even greater critical thinking! AI is only an enabler but not a solutions tool in and of itself.

AI is part of Futures thinking toolkit; FT is a way to break-free of conventional thinking, and AI can assist here with e.g. scenarios generation

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Greater need for futures thinking across the whole planning and development domain; FT is under-represented and there is a lack of understanding of what it is and how it can be used to improve decision-making especially over 'post-project/post-policy timeframes to achieve inter-generational equity

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Create greater awareness and familiarisation with Futures Thinking tools, and to utilise them in all IA processes. This extends way beyond project-based decisions and timelines; most especially many practitioners face difficulties in thinking beyond project -level scale, whilst others think at mega-scales too broad to be useful at individual project scale.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

1. Futures Thinking training courses for IA professionals, decision-makers etc. are desperately needed. 'How to' best practice of applying FT tools like scenarios development are needed. 2. Move away for scaled-down, tick-box sustainability at project-level to achieve holistic 'sustainabilty' as per the deep understanding and principles within Indigenous knowledges

SESSION NAME

Assessment of Climate-Related Disasters and Adaptation Actions

NAME(S) OF SESSION CHAIR(S):

Emdad Haque

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI can help climate impact assessment in 3 areas: climate risk analysis, context-aware solutions, and technical and logical feasibility assessment. AI can be useful for predictive analytics, natural language processing, data integration, and scenario modeling.

In the context of climate-related uncertainty, AI can help to improve the degree of confidence in predicting climatic scenarios and atmospheric behaviour.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Yes, serious attention should be kept on the successful integration of scientific/technical dimensions with societal dimensions/variables in impact assessment efforts.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

The aspects of multidimensionality in addressing climate impact assessment need both academic and policy attentions. Social and climate justice perspectives were recommended by the session presenters.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

SESSION NAME

Evolving the state of the art in Psycho-social impact assessment

NAME(S) OF SESSION CHAIR(S):

Helen Ross

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI requires monumental new amounts of energy, requiring massive new power generation and transmission facilities, with climate change impacts and impacts on environments and people. A comparative test of AI versus human analysis of open-ended survey responses showed AI quite effective with the qualitative data but making basic errors with the quantitative.

Raphael Carvalho study of data from public consultations on a Canadian electricity substation showed some value in data analysis, with caution needed. Policy attention is needed on the electricity requirements of AI processing: this has major, neglected, requirements running counter to meeting climate targets. (Michael Edelstein paper)

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Discussion focused on how participants would like the state of the art in psycho-social impact assessment to evolve. Key points were: - understanding people's histories, perceptual sets, leading to the need for healing and repair. - desire for PSIA and HIA to come together more; both are evolving (like to see more sociology and anthropology in HIA too) - examining the psychology of decision-makers, e.g. biases leading community concerns to be dismissed - more time to document; allowing time for people to heal - PSIA is a progressive idea, it should be in every study. It can identify the seeds of distrust. - natural events can have psychological impacts. - need a more clear gender and intersectionality focus - cultural impacts do not only occur in the 'global south'. e.g. windfarms affecting culturally important peatlands in Scotland - value of bringing in conflict resolution and skills, help people to listen to each other

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Consider the potential psycho-social dimensions in your study context. Seek to incorporate PSIA (often with or within SIA, HIA) if psychological dimensions are likely. Recognise that environmental change can be culturally meaningful, hence the psycho-social (and social, and health) can be very directly related. Approach engagement processes in awareness of potential PSI backgrounds, carrying hurt and mistrust, cumulative impacts, forward.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

As above: - Consider the potential psycho-social dimensions in change contexts. Seek to understand these, and proceed sensitively - Recognise cultural dimensions and PSIs can arise with environmental change, so environmental impacts cannot be considered as separate - Approach engagement processes in awareness of potential PSI backgrounds, carrying hurt and mistrust, cumulative impacts, forward. These may apply to governments, international agencies (including in development, and disaster risk management), planners (e.g. using SEA), project proponents

SESSION NAME

Updating Best Practice Principles for Engaging with Indigenous Peoples

NAME(S) OF SESSION CHAIR(S):

Diana Lewis, Dyanna Jolly

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

We had a brief discussion about using AI to develop best practice principles, and the ability of AI to collate and summarise the information already available on best practice.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

There is an urgent need to update the best practice principles for engaging with Indigenous Peoples, to align with current examples of best practice. Many Indigenous nations, groups, communities and organisations have developed their own road maps. The IAIA best practice principles can provide high level guidance while recognising and providing for diversity of Indigenous contexts, realities and experiences. In this session we were asked to progress the update as soon as possible.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

SESSION NAME

Blue Economy, SEA and Marine Spatial Planning

NAME(S) OF SESSION CHAIR(S):

Arthur Neher, Alexandre Puisais

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

- Although AI was not a central theme in the session, one presentation briefly explored the potential of AI to enhance environmental modelling for marine spatial planning. This highlighted a growing interest in how AI tools might improve the quality and efficiency of assessments. - A broader takeaway from the session was the recognition that AI is not just a neutral tool to be applied in impact assessment, it is also its physical embeddedness. The energy-intensive infrastructure required to run AI models, carries its own environmental impacts on marine resources. Ranging from decarbonizing data center energy use (e.g. offshore wind or ocean cooling systems) to evaluating the impacts of deep-sea mining for resources used in AI hardware, or submerging data centers for cooling.

As AI was not the main topic, no specific example was given.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Strategic Environmental Assessment can support Marine Spatial Planning into integrating the environmental and social elements into planning and decision-making for the marine and maritime environment. It requires, equally to terrestrial spatial planning much topic information that is substantially corroborated.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Adopt a systems perspective when considering AI: Practitioners should not only explore how AI can support and enhance impact assessment—such as through improved environmental modelling for marine spatial planning—but also recognize that the use of AI generates new dependencies on marine and coastal resources. The infrastructure required to run AI models and the demand for critical minerals potentially sourced from the deep-sea mining mean that marine environments are increasingly implicated in the digital transition. This reinforces the strategic importance of marine spatial planning as a tool to balance competing demands on marine ecosystems, including those driven by the growth of AI technologies.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Sectoral interventions in the marine environment need careful consideration of the natural and social environment, i.e. in environmentally vulnerable and sensitive areas, as well as when many past, present and planned activities occupy the same marine space. At the same time, cumulative impact assessment, transboundary impacts and activities impacting the areas beyond national jurisdiction (e.g. mining of critical minerals in the deep seas) need to be considered and taken-up in national (marine) planning supported with integrated SEA.

SESSION NAME

Transparency and Data Sharing in Impact Assessment for the Energy Transition

NAME(S) OF SESSION CHAIR(S):

Diana Arbelaez Ruiz, Rauno Sairinen

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

- The risk of hallucination is a concern for practitioners - There are ways to mitigate the risk of hallucination by restricting LLM to using data within specified datasets. - Teams that are interdisciplinary and collaborative can help create the depth of expertise on IA and IA for effective solutions. - Data standardization has an important role in enabling the use of AI - Data integrity is fundamental to good AI applications in IA - AI applications need to be inclusive - AI, if well used, can democratize impact assessment - AI, if well used, can provide more opportunities to triangulate data

- Denmark's EA Tools and EA Hub that Dorthe Holme presented (impressive example of transparency, ambition, and cost-effectiveness) These tools can be useful for policy makers, First Nations, community members, academics and project proponents.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Policy responses need to keep up with the data needs of impact assessment. For example, policy should require full disclosure of existing impact assessment data in accessible formats following FAIR standards. We face similar questions in impact assessment at various scales, yet policies, guidance and procedures at those scales - e.g. municipal, regional, national - can vary widely. Often procedures or policies are not keeping up with current data needs. These policies/procedures can be at odds with each other, or not evaluated for their effectiveness.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

We need to learn about AI as much as we can and help educate younger practitioners and researchers in the intersection between AI and AI. AI applications are diverse and we need to understand what applications are suitable for what purposes so that it enhances our capabilities and improves results and processes, and we can mitigate the risk of exclusion or misrepresentation.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Policy needs to foster open data sharing. Policy needs to foster AI applications that are inclusive and that are designed with the participation of people from different backgrounds.

SESSION NAME

Linking Global Commons Issues to IA: Fit for Future Actions and Next Steps

NAME(S) OF SESSION CHAIR(S):

Westen Fisher and Barry Wiesner

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The Initiative is exploring new strategies, professional development programs, including potential AI integration to drive earth and human systems change.

As above, and training in the use of AI is required.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Six of nine planetary boundaries have been breached, underscoring the gaps in our current impact assessment methods and tools.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

We have a current focus on developing alternative IA training incorporating earth and human systems thinking on the African continent, that can be replicated in other world regions.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

See the bigger picture. Planetary boundaries and earth and human systems need to shape how we govern and determine our policies. We need start with ourselves.

SESSION NAME

The Biodiversity Significance of Some Modified Habitats: Implications for Impact Assessment (Parts I and II)

NAME(S) OF SESSION CHAIR(S):

George Ledec

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI will greatly increase the global demand for electricity, thereby making even more challenging the achievement of a low-carbon energy transition in ways that are biodiversity-friendly and socially just.

AI expansion will require the construction of new data centers, some of which may be proposed in areas of modified or natural habitats with significant biodiversity value.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Some modified habitats are important for biodiversity—some even qualify as critical habitats. However, areas labeled as “modified habitats” are often neglected or overlooked with respect to their actual (or, with restoration, potential) biodiversity significance.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

IA practitioners should adequately assess the biodiversity value of each project area, even if it is classified as one or more modified habitats.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Recommendation to government officials and project developers: Not all modified habitats are well suited for new infrastructure, housing construction, energy development, or agricultural expansion projects—some already have high biodiversity value, some have potential for ecological restoration, and some may be valuable as protected open space for recreation and other public uses.

SESSION NAME

Beyond the Buzz: Real-World AI Applications in IA

NAME(S) OF SESSION CHAIR(S):

Andy W L Chung, Ana Maria Quintero

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

To summarize the key takeaways using the IAIA acronym: I for INNOVATE: Technology is available; the real impact lies in how we innovate to apply it effectively in different contexts. A for AI READY: Data is king, but datasets must be organized, cleaned, and standardized to be truly AI ready. I for INTEGRATE: Impact assessments often rely on multiple data streams with different structures and histories; success depends on finding ways to integrate them seamlessly. A for ADAPTIVE: There are no golden rules yet—impact assessment practitioners must remain adaptive, evolving their approaches alongside rapid technological change.

OPPORTUNITIES:- AI offers powerful tools to inform policies, shape strategic planning, and optimize project execution. By integrating diverse data streams—such as satellite imagery, drone footage, and field measurements—AI enables deeper environmental insights and real-time monitoring (Arik, Stefansson). This allows policymakers and project leaders to anticipate risks, design more resilient strategies, and implement projects with greater precision and efficiency (Carter, Ziduche). AI also fosters innovation by pushing impact assessment toward adaptive, data-driven approaches that evolve alongside changing environmental and social conditions (Scott-Brown). CHALLENGES:- Data quality, standardization, and interoperability remain major hurdles (Stefansson, Carter); without clean, structured data, AI outputs risk being unreliable or misaligned with policy goals. Presenters stressed that while AI can analyze and recommend, final decisions must stay human-led to uphold accountability and ethics (Ziduche). Building trust and transparency around AI processes is critical to gain public and stakeholder confidence (Arik, Scott-Brown), and tools must be flexible enough to adapt to local legal, environmental, and cultural contexts (Carter).

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

AI is here to stay; there’s no turning back. The session highlighted the critical importance of data governance and cross-disciplinary collaboration, calling for stronger data management practices and clearer standards. It also reinforced that community and stakeholder engagement must remain central, and emphasized the need for capacity-building in both digital literacy and environmental expertise.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Impact assessment has always been teamwork—now AI is part of the team. Practitioners should treat AI as a valuable team member: build trust in its outputs, integrate it thoughtfully, and leverage its strengths to create added value for clients. The goal is not just to use AI, but to co-create value with clients—by co-building models, tailoring solutions together, and strengthening clients’ capacity to engage with AI tools themselves.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

For project proponents, clients, and project owners, the key recommendation is to see AI as a new team member within your project delivery ecosystem. Invest time in understanding AI’s capabilities and limitations, build trust in its outputs, and work closely with practitioners to integrate AI effectively. The goal is not simply to deploy AI, but to co-create value—by co-developing models, customizing AI solutions to your project’s unique needs, and building in-house capacity to engage confidently with AI tools.

SESSION NAME

IAIA Climate Change Action Plan: What’s New? What’s Next?

NAME(S) OF SESSION CHAIR(S):

Pantea Niksirat

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Integrating AI modeling into impact assessment (IA), and developing tools to optimize sustainable taxonomy, enhance legal instruments, and improve practices and legislation.

Network analysis and country-specific examples, including Nigeria, South Africa, Australia, Myanmar, United States, and Uganda.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Four themes were discussed: Sustainable Finance and Impact Assessment (IA); IA Legal Procedures and Institutional Frameworks; Tiering IA from Global to Project Level to Address Climate Change; and Emerging Topics and New Ideas.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Improved coordination between IA practitioners and financial and governmental institutions, along with better integration of AI models to identify procedural gaps in climate change mitigation and adaptation.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Climate change action must emphasize financial impacts, as funding is crucial—particularly in developing countries—for effective mitigation and adaptation. Involving lawyers early in the planning process is essential, with a greater role needed for environmental lawyers in shaping climate change strategies.

SESSION NAME

400 Pivoting Policies: Evidence-Based Environmental Policymaking and its Impact on Renewable Energy Development

NAME(S) OF SESSION CHAIR(S):

Jana Minnaar

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

N/A

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The effectiveness and impact of environmental policy is highly context-dependent, influenced by existing energy infrastructure, political will, economic incentives, and social priorities—as shown by the contrasting experiences of Canada, Iran and South Africa.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

As seen in Alberta (Canada) and Iran, well-intended or unrelated policies (e.g., fossil fuel subsidies or approval pauses) can have negative spillover effects. Impact practitioners should flag such risks clearly in their assessments.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Recommendations for national policy makers are to ensure that energy, environmental, and economic policies work together rather than at cross-purposes (e.g., removing fossil fuel subsidies while promoting renewable investment).

SESSION NAME

AI in the USA: Technology’s Potential to Improve the National Environmental Policy Act (NEPA) IA Process

NAME(S) OF SESSION CHAIR(S):

Michael Smith

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The incorporation of AI in the U.S. National Environmental Policy Act (NEPA) process is still in a very nascent state. Where it is being used, it is mostly in relation to organizing information and data collection, and categorizing public comments. Many concerns were expressed about issues with accuracy in using AI for more advanced tasks in the impact assessment process.

There has been a useful effort to use AI for data scraping to provide information from existing impact assessment documents that are not widely available in any one location, and also efforts to use AI to analyze the need to update old/stale data on environmental factors.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

There was a lot of discussion in the session on the recent changes to the practice of NEPA that have occurred in the past three months through the actions of the new presidential administration, and what the implications of those changes would be for the practice of impact assessment. The conclusion from the speakers in the session is the jury is still out whether the changes may be temporary and changed by a future administration, or whether the practice of impact assessment in the U.S. will suffer more long-lasting or even permanent damage.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

1). Explore the potential of AI to be useful in the impact assessment process, as there is solid potential 2). Be cautious in implementing AI methods in the impact assessment process, and when using - always have humans fact check information and results developed through the use of AI.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

The recent changes to how the NEPA process is administered require patience for those non-federal stakeholders requiring EIA review to gain approval for their projects, as the chaos and uncertainty many of the changes have introduced have led to a slowdown in project reviews and approvals.

SESSION NAME

IAIA and Cumulative Effects: A new Section and a Busy Future

NAME(S) OF SESSION CHAIR(S):

Martin Haefele, Johanna Gordon

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Cumulative effects assessment has been a key concept for a long time and it's importance is growing. It's implementation has been lacking and continues to do so. There is cautious hope that AI can assist in analyzing and bringing together vast historical data sets. There does not appear to be successful examples yet.

n/a

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Cumulative effects assessment has been a key concept for a long time, it's importance is growing, and in recent year's It's applications, though in many respects still anemic has made significant progress. That said, even the experts in the field feel that there is not yet a good and commonly accepted definition.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Join the Cumulative Effects Section!

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

No recommendations yet, but we identified an urgent need for them.

SESSION NAME

Strategic Thinking in Impact Assessment to enhance sustainability (I and II)

NAME(S) OF SESSION CHAIR(S):

Maria Partidario

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

The importance of flexibility, broad and long-term thinking and the general use of strategic thinking in the assessment; and of multi-actor collaboration in discussing how to identify and understand the problem

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Be flexible, strategic, collaborative in your thinking and practice

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Be flexible, strategic, collaborative in your thinking and practice

SESSION NAME

Can AI Be an Effective Tool for Supporting Country Biodiversity and Forestry Management?

NAME(S) OF SESSION CHAIR(S):

Ahmed Sanda

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The most important insights were: (1) The audience was looking new ways (fresh) of doing things which has human component(s). (2) They also want to interact more trusted in spaces.

The presenters unanimously agreed and a substantial number in the audience believe that AI provides great opportunities particularly in monitoring and some aspects of analysis. This would potentially reduce time taken to monitor and analyze huge amount of data. The challenges it faces in shaping policies are: the lack of regulations and that the information about its capabilities is confusing. Strategies: Developing strategies while relying on old protocol and or regulation are vey difficult; inadequate and or untrusted data. Project: General reluctance to use AI data; not enough data from some parts of the world- thus making analysis difficult if not impossible; it needs more education because many ask the wrong question as such AI is not its optimal.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

NON

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

(1) No One Single Person Knows it All, But Our Collective Knowledge May Just be Enough. (2) Increase your curiosity about your local environment beyond your field of expertise and your interaction with communities outside your environment or comfort zone.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Policy makers: hasten the replacement of old protocol with new ones. Experts: fill in the data gaps in the international repository 'banks'. Knowledge bearers: educate and inform the public as much as humanly possible.

SESSION NAME

Can AI Support Long-Term Monitoring and Adaptive Management?

NAME(S) OF SESSION CHAIR(S):

Janet Blackadar

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI can vastly improve processing time for data sets collected over extended space and time such as the ocean. There are many possible uses for AI in both monitoring of species and adaptive management of same. Large language models (LLMs) can be very useful in data mining and interpretation of paper based data.

One presentation focussed on the use of AI to assist in processing data from camera traps for species at risk in remote locations thus reducing the Health and Safety risk of having people in the field for extended periods and increasing the likelihood of capturing occurrences of "shy" species.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

It is important that potential bias in AI training be recognized and that the privacy of data be kept particularly as it relates to species at risk

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

AI is very helpful for processing data and allowing more monitoring over large spatial extents, but there is still great potential for bias in the model and these biases and limitations of the programs need to be understood and explicit if we are to use the outputs for decision making.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

For regulatory decision makers and policy makers alike AI is very helpful for processing data and allowing more monitoring over large spatial extents, but there is still great potential for bias in the model and these biases and limitations of the programs need to be understood and explicit if we are to use the outputs for decision making.

SESSION NAME

Human Rights Impact Assessment of Digital Technologies

NAME(S) OF SESSION CHAIR(S):

Andy Symington, Gabriela Factor

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

There are two important points regarding the intersection between digital technologies including AI and human rights. First, digital technologies can support human rights impact assessment (HRIA). Second, digital technologies can pose serious risks to human rights enjoyment. Therefore digital technologies, including when used in the context of ESIAs and HRIAs, should be assessed from a human rights perspective. Some of the salient human rights risks include right to privacy, data protection and non-discrimination, right to freedom of expression, as result of data bias, unequal participation in the development of technologies, and lack of consent for data processing and use.

One presentation focused on the use of AI and IT tools in human rights management. For example, one in-house solution discussed was that of a human rights data lake that is a centralised repository to store, process and secure large amounts of structures and unstructured data from internal systems (grievance, HR, procurement, audits, legal processes).

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

It is essential that users of digital technologies in Impact Assessment to do a due diligence, assess and address the risk to human right of such technologies, for example, profiling, data leaks, reinforcing biases, discrimination, etc.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

2 recommendations: 1. Mature emerging practices of using digital tools in HRIAs and ESIAs, but pay heightened attention to potential human rights risks. 2. Raise awareness on the need to conduct human rights impact assessments for the development and deployment of digital technologies.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Recommendations for policy makers, ESIA and HRIA practitioners, and the tech sector. 1. Mature emerging practices of using digital tools in HRIAs and ESIAs, but pay heightened attention to potential human rights risks. 2. Raise awareness on the need to conduct human rights impact assessments for the development and deployment of digital technologies.

SESSION NAME

Emerging AI applications in IA (I and II)

NAME(S) OF SESSION CHAIR(S):

Alan Bond

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI is a fantastic resource that can increase efficiency and also accuracy, but needs to be used with caution and checked. Integrating AI into a business is a long term task that requires a strategy. AI can help to remove biases that may be inherent in human judgements.

All talks focused on specific examples of the use of AI in IA. So there were examples related to use of AI for data correction; risk assessment; the analysis of the full range of tasks where AI could benefit IA practice; gap analysis; and development of better focused environmental and social management plans.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Keep thinking about how AI can help to improve efficiency and or accuracy, but never lose sight of the need for accountability forn AI outputs (which reside with human beings).

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Accountability is key - decision makers in the room were very clear that they need to be able to rely on evidence presented to them.

SESSION NAME

Asian S3EA: Strategic, Spatial and Sustainable EA in AI Era I and II

NAME(S) OF SESSION CHAIR(S):

Kii Hayashi, Myungjin Kim

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI as an analysis tool was conducted in several presentations and showed the usefulness of it’s application to an assessment.

AI is used as a platform for analysis. And several data-driven methods were used also in GIS study.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Several new practical examples and studies were conducted and reported in Asian region.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Currently, AI can be a basic platform for analysis and be used for many studies. This tendency can be continued. Then comprehensive AI will be a basic analysis platform an IA study in the near future.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Analysis platform supported by AI tool can be expanded. Then application of this platform can be useful for an policy decisions.

SESSION NAME

Towards Improving Social Impact Management in the Mining Sector

NAME(S) OF SESSION CHAIR(S):

Kerry Fairley

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

AI was not mentioned in the session

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

AI was not mentioned in the session.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

A key topic raised and discussed in the session was the consideration of intangible social impacts in E(S)IA processes. Intangible social impacts should not be equated to immeasurable impacts as there are methods for quantitatively assessing such intangible impacts. These methods need to be explored and developed further.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Intangible social impacts are only being considered if required to do so - mainly as a result of standards of lender institutions (such as IFC Performance Standards) that require for these to be done. The consideration of intangible social impacts need to be written into policy or legislative requirements to ensure that these are considered in the E(S)IA process.

SESSION NAME

Using AI Technologies to Quantify Environmental Impacts on Complex Systems

NAME(S) OF SESSION CHAIR(S):

Zishu Wang, Chunyan Wang

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

While AI technology demonstrates promising potential in enabling quantitative impact assessment of complex systems, the current methodological framework remains nascent, presenting substantial challenges in standardization and scalability.

a. An Indicator System for Assessing Intelligent Environmental Governance; b. AI-Enhanced Turbidity Prediction: Mitigating Impacts on Dam Reservoirs; c. A framework for basin-scale water ecological assessment based on multi-source data.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Participants examined China's eco-environmental zoning regulation policies and their implementation efficacy during the session, with particular emphasis on data-driven governance frameworks and cross-regional coordination mechanisms.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

AI technologies are advancing environmental assessment methodologies to bolster system-level decision-making processes within integrated socio-economic-environmental systems.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

For governmental decision-makers, the application of AI technologies should currently remain at the decision-support stage. Assessment outcomes require holistic integration with local socioeconomic contexts, while systematically accounting for uncertainty factors embedded in policymaking cycles.

SESSION NAME

Reform of EsIA: the Good, the Bad and the Ugly

NAME(S) OF SESSION CHAIR(S):

Cheryl E Wasserman

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

1- AI has the potential to create efficiencies in IA processes including intergovernmental coordination, screening, scoping and mitigation as well as public engagement. 2- AI should be coupled with Information technology and GIS tools to support the entire governance and implementation process. That process includes follow up compliance monitoring and enforcement. 3- AI is a complement, not the answer to the search for reduced time, cost and increased quality of the EsIA process in meeting its goals.

1- Opportunity presented by direct reporting of monitoring data thru AI for regulatory analysis and follow up using AI within the company for its own compliance management. 2-Opportunity to Interpret monitoring data and performance for the public. 3-Opportunity to translate EsIA analysis into plain language and alternative languages. 4- Opportunity to link comments to how they are addressed. 5- Opportunity to assure integrity in linking impacts to mitigation plans.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

1- Time limits are counter productive because they can compromise quality, lead to disputes and people find a way around them. 2- Data quality, transparency and easy access are imperative. 3- Public engagement after project approval helps support compliance. 4- Impediment in areas without internet service. 5- Significant concerns raised about some proposed government reforms to assign aspects of EsIA assessment and review to entities with specialized expertise e.g. biodiversity; ignoring climate change and environmental justice considerations.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

1- Be open to explore possibilities but be aware of misuse and abuse of AI and the fact that it will not replace human responsibility. 2-Difficulty of finding errors or lack of depth in nicely written AI generated documents. 3-Start small with control over substance and control. 4-Just as metadata can provide information on data use and quality for specific purposes, need to identify sources of modeled impact information.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

1-Donors and countries: The investment in implementation systems using both IT and GIS and introduction of AI will be costly and a collaborative effort should be initiated with donors to create a flexible digital platform that countries can use to introduce their own data, standards, policies without repeating investments in individual countries. 2-Practitioners and project proponents: there are no shortcuts, only efficiencies, but need to focus on outcomes and results.

SESSION NAME

Artificial Intelligence Tools to Improve Impact Assessment of Linear Infrastructure Project

NAME(S) OF SESSION CHAIR(S):

Franck Duchassin

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The diversity of the 4 proposed application of AI to improve or ease the faster creation, accuracy, quality of the IA.

Virtual reality with AI permits a better comprehension of the project. Translation with AI improves the communications.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Despite the great potential of AI, human judgement must remain in the process of IA.. AI must be used only as a tool.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Be mindful of the risks of utilizing AI and use it for the communities, not against them.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

N/A

SESSION NAME

The Emerging Role of Artificial Intelligence in the Future of Impact Assessment

NAME(S) OF SESSION CHAIR(S):

ALexandra Jiricka-Pürrer; Gesa Geißler

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

- need for clarifiying what we are talking about when talking about AI, clarifying definitions - huge interest and perceived relevance of AI in IA - strong need for training and capacity building on AI - going beyond LLM application to ML and deep learning with data centered approaches

- example from Denmark, Danish Environmental Hub - Austrian workshops on renewable energy planning and potential for AI supporting this

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

- relevance of good quality data - need for multi-disciplinary teams for IA

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

- dive deeper into the topic of AI in IA in multidisciplinary teams which combine thematic environmental and social expertise with data science and AI

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

- be open for benefits of novel AI approaches to support qualitative data based decision making - revising standards and guidance for accounting of AI

SESSION NAME

The Validation of Impact Assessments, Adaptive Management, and Reassessments

NAME(S) OF SESSION CHAIR(S):

Daniel Stuckless, Margaret Luker

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

The availability of instant data and data confidence using calibrated AI and AI supported tools/equipment (ie cameras for biodiversity or audio recognition devices. The simplification of tremendous amounts of information readily available support positions or conclusions.

The use of proprietary 'chat bots' or internal AI systems for large or global institutions vs those that are publicly available. Example, growing an internal World Bank chat bot that is independent of other AI internal centers.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Not a lesson but an observation, younger professionals are more optimistic with the use of AI in their work and from the questions, order traditional professionals are skeptic and prefer the certainty of 'doing it their own way'.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Be open open minded and adapt. It's going to change fast and it will be tough to keep up.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Generally speaking with Impact Assessment, the human element provides the exceptions based on weighted preferences or previous knowledge. It may be necessary for guidance parameters to be generally applied to AI based recommendations or summaries.

SESSION NAME

What is needed to achieve trusted and efficient IA for 'green' projects

NAME(S) OF SESSION CHAIR(S):

Celesa Horvath

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

No discussion of AI.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

IA practitioners and participants (including policy and decision makers) must be clear about what they mean when they refer to "streamlining" and improving "efficiency" and "effectiveness" in the context of IA to advance dialogue. "Streamlining" IA cannot be focused only on shortening timelines; it must maintain trust in the process and its outcomes. Time is not the only important metric in the IA process. Efficiency initiatives cannot focus only on the IA phase; must consider trust in the entire project life cycle. We need system-wide integration: IA is not the only means by which issues can or should be addressed; IA will be more efficient if issues are dealt with in the appropriate part of the system (e.g., strategic planning, land use planning, permitting, monitoring, etc.). Need to develop stronger competencies in decision-making skills (e.g., knowing when there is enough information to inform a decision, being able to say 'no', to filter comments, to maintain focused scope). Need to reconcile standardization and flexibility.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Demonstrating clarity, concision, and reasoning in IA documentation can build trust, improve efficiency and quality of decision-making. Engage early, build relationships. Need to demonstrate trustworthiness not just in IA phase but beyond, throughout project lifecycle (e.g., demonstrating success in adaptive management).

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Explore opportunities to improve efficiency, but be transparent about what these involve, how these will work, and what the consequences may be. Opportunities exist to standardize conditions to enable well understood effects and proven effective mitigation to be removed from the scope of assessment (but still addressed in the regulatory regime); build a credible and trusted repository of information. Articulate principles to guide efficiency improvement initiatives.

SESSION NAME

Improving Governance and Decision Making in IA: Addressing Ecological and Participatory Challenges

NAME(S) OF SESSION CHAIR(S):

Jane McQuitty

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Three initial presentations focused on the integration of relatively new monitoring technologies—two on the use of aerial monitoring for avi-fauna and marine mammals. Presenters saw the technologies as requiring a period of human interpretive oversight, after which point the curated data could form an archive—potentially evolving into a data bank suitable for machine learning.

n/a We addressed persistent challenges in data collection and, in my case, the hegemonic filters that shape what counts as significant. In the Canadian context, the core issue was not merely how data was interpreted, but whether it could legitimately represent the environmentalism of the impacted group—specifically, whether their concerns could be made intelligible to decision makers, given that the data was used to uphold a form of environmentalism disconnected from their actual concerns with environmental justice

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

Key lessons emerged around the promise and politics of new monitoring technologies. Both eDNA and aerial survey techniques proved highly effective in challenging environments. The two presentations on eDNA highlighted its speed, cost efficiency, and non-intrusive nature, making clear its value for aquatic monitoring. The aerial survey presentation showcased an ingenious approach to minimizing the impact of surveillance in difficult offshore conditions, establishing a useful baseline for assessing how wind turbines affect wildlife. The final presentation underscored a crucial point: in deliberative democratic contexts, Environmental Impact Assessment (EIA) reporting carries demonstrable political weight. Decision makers, the public, and practitioners must all recognize where data collection ends and political interpretation begins. EIA practitioners cannot see themselves merely as neutral data collectors, just as the public and decision makers must avoid treating EIA outputs as the voice of pure science. Ultimately, it is not nature that interprets the data—but people, with interests, values, and assumptions. Acknowledging this is essential to making informed, democratic decisions.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

Adopt eDNA for finding baselines and monitoring impact to those baselines in aquatic settings, aerial survey for a baseline from which to monitor marine mammals and birds seems promising. For assessment of the legitimacy of the performance of setting guide rails for decision makers carried out by the airing of EIA highlights at public fora, in countries like Canada where there is no Arhus agreement covering all the nation's populations, consider that all actors in democratic deliberative fora should have access to EIA reports and 30 to 60 days response time.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Adopt eDNA for establishing baselines and monitoring changes in aquatic environments; use aerial surveys. Aerial surveys show promise for establishing baselines to monitor marine mammals and birds. When it comes to evaluating the legitimacy of Environmental Impact Assessments (EIA) as tools for guiding decision makers—particularly through the public presentation of EIA highlights—countries like Canada, which lack full Aarhus Convention-style protections, should ensure that all participants in democratic deliberative fora have full access to EIA reports and are granted 30 to 60 days to respond. Developer-hired EIA professionals at these hearings should volunteer the place where their report transitions from accumulation of data to interpretation of the meaning of that data in terms of a clearly stated social priority "X".

SESSION NAME

Applying and strengthening qualitative methods for effective IA

NAME(S) OF SESSION CHAIR(S):

John Sinclair

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

- potential use if AI in the creation of summaries of qualitative data - identification of themes. - there was not a lot of comment on AI

- one challenge noted is that AI could be (probably is) used to generate qualitative data without actually collecting that data in the field.

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

- that more attention needs to be paid to qualitative methods in IA especially since many IA decisions rest of interpretations of people's current situation. - indigenous methods need to be used. - barriers to the use of qualitative methods, such as the sense that not lacks rigour, needs to be addressed.

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

- the is need for more training in the use of qualitative methods - quantitative researchers should not be doing the qualitative work. - spend time with Indigenous communities to build relations and collect data collaboratively. - expand the tool kit of qualitative methods being used.

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

- ensure terms of reference for IA's include indicators that are best supported by qualitative analysis.

SESSION NAME

The good the bad the ugly: the role of AI in public participation in IA

NAME(S) OF SESSION CHAIR(S):

Tim Peirson-Smith

1) IF PRESENTATIONS OR DISCUSSION IN YOUR SESSION TOUCHED ON THE USE OF AI IN IA, TWO KEY MESSAGES ON THE THEME "IMPACT ASSESSMENT IN THE AGE OF ARTIFICIAL INTELLIGENCE"

1A. WHAT WERE THE MOST IMPORTANT INSIGHTS OR TAKEAWAYS FROM YOUR SESSION REGARDING THE INTERSECTION OF IMPACT ASSESSMENT AND ARTIFICIAL INTELLIGENCE (AI)?

Keep Humans in the PP loop Ai is merely a tool in PP and must be checked

As above

2. ARE THERE ANY ADDITIONAL KEY MESSAGES OR LESSONS FROM YOUR SESSION THAT ARE RELEVANT TO IMPACT ASSESSMENT MORE BROADLY–BEYOND THE THEME OF AI?

As above

3. WHAT RECOMMENDATIONS WOULD YOU OFFER TO IMPACT ASSESSMENT PRACTITIONERS BASED ON THE DISCUSSIONS IN YOUR SESSION?

As above

4. WHAT RECOMMENDATIONS EMERGED FROM YOUR SESSION FOR POLICYMAKERS AND OTHER KEY STAKEHOLDERS? (PLEASE SPECIFY THE TYPE OF STAKEHOLDER, WHERE POSSIBLE.)

Session Chair Reports

Impact Assessment in the Age of Artificial Intelligence

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

SESSION NAME

DOWNLOADS

FOLLOW US