Session
Organizer 1: Intergovernmental Organization, Western European and Others Group (WEOG)
Speaker 1: Lillian Røstad, Technical Community, Western European and Others Group (WEOG)
Speaker 2: Pettersen Klas Henning, Government, Western European and Others Group (WEOG)
Speaker 3: Signe Riemer-Sørensen, Technical Community, Western European and Others Group (WEOG)
Speaker 4: Riegler Michael A., Intergovernmental Organization, Western European and Others Group (WEOG)
Speaker 2: Pettersen Klas Henning, Government, Western European and Others Group (WEOG)
Speaker 3: Signe Riemer-Sørensen, Technical Community, Western European and Others Group (WEOG)
Speaker 4: Riegler Michael A., Intergovernmental Organization, Western European and Others Group (WEOG)
Format
Roundtable
Duration (minutes): 90
Format description: The roundtable format is ideal for exploring AI agents' governance implications as it creates a non-hierarchical environment where diverse stakeholders can engage in meaningful dialogue. Unlike a panel, where experts present to a passive audience, a roundtable encourages active participation from all attendees, fostering the exchange of diverse perspectives from technologists, policymakers, civil society, and business representatives. The circular seating arrangement visually reinforces that all participants' contributions are equally valued, breaking down barriers between "experts" and "audience." This approach aligns perfectly with our goal of raising awareness and kickstarting discussions about autonomous AI agents as emergent actors in digital governance. The 90-minute duration provides sufficient time to explore complex governance questions while maintaining engagement - allowing for framing key issues, guided discussion on policy questions, collaborative identification of governance principles, and synthesizing concrete next steps for ongoing collaboration.
Duration (minutes): 90
Format description: The roundtable format is ideal for exploring AI agents' governance implications as it creates a non-hierarchical environment where diverse stakeholders can engage in meaningful dialogue. Unlike a panel, where experts present to a passive audience, a roundtable encourages active participation from all attendees, fostering the exchange of diverse perspectives from technologists, policymakers, civil society, and business representatives. The circular seating arrangement visually reinforces that all participants' contributions are equally valued, breaking down barriers between "experts" and "audience." This approach aligns perfectly with our goal of raising awareness and kickstarting discussions about autonomous AI agents as emergent actors in digital governance. The 90-minute duration provides sufficient time to explore complex governance questions while maintaining engagement - allowing for framing key issues, guided discussion on policy questions, collaborative identification of governance principles, and synthesizing concrete next steps for ongoing collaboration.
Policy Question(s)
A. How should accountability frameworks evolve to address autonomous AI agents that make decisions with limited human oversight?
B. What technical standards and transparency requirements should govern AI agents' interactions across the digital ecosystem?
C. How can multistakeholder governance approaches ensure AI agents develop in ways that respect human rights while fostering innovation?
What will participants gain from attending this session? -A deeper understanding of how autonomous AI agents are reshaping digital interactions and governance landscapes
Insights into emerging technical standards and ethical frameworks for responsible AI agent deployment
-Knowledge of potential regulatory approaches from diverse global perspectives
-Practical considerations for addressing accountability, transparency, and equity issues
-Networking opportunities with cross-sector stakeholders addressing similar challenges
-Concrete principles that could inform organizational strategies and policy development for AI agent governance
Description:
In the rapidly evolving digital landscape, AI agents are emerging as new actors that operate autonomously on behalf of users, organizations, and platforms. These agents—ranging from personal assistants to sophisticated autonomous systems—are increasingly mediating our interactions with digital spaces, making decisions, processing information, and executing tasks with limited human oversight. This session explores the governance implications of this shift, examining how existing frameworks must adapt to account for these non-human participants in our digital ecosystem. We will investigate critical questions including: How do autonomous AI agents challenge traditional notions of accountability and transparency? What technical standards and ethical guidelines should govern their deployment? How might these agents exacerbate or potentially bridge digital divides? And what role can multistakeholder governance play in ensuring these technologies develop in ways that respect human rights while fostering innovation? Through interactive dialogue between technologists, policymakers, civil society, and business stakeholders, this session aims to identify concrete principles for responsible AI agent governance that balance innovation with necessary safeguards for a trustworthy digital future.
In the rapidly evolving digital landscape, AI agents are emerging as new actors that operate autonomously on behalf of users, organizations, and platforms. These agents—ranging from personal assistants to sophisticated autonomous systems—are increasingly mediating our interactions with digital spaces, making decisions, processing information, and executing tasks with limited human oversight. This session explores the governance implications of this shift, examining how existing frameworks must adapt to account for these non-human participants in our digital ecosystem. We will investigate critical questions including: How do autonomous AI agents challenge traditional notions of accountability and transparency? What technical standards and ethical guidelines should govern their deployment? How might these agents exacerbate or potentially bridge digital divides? And what role can multistakeholder governance play in ensuring these technologies develop in ways that respect human rights while fostering innovation? Through interactive dialogue between technologists, policymakers, civil society, and business stakeholders, this session aims to identify concrete principles for responsible AI agent governance that balance innovation with necessary safeguards for a trustworthy digital future.
Expected Outcomes
-Increased awareness among IGF stakeholders about the emerging governance challenges posed by autonomous AI agents
-A published set of draft principles for responsible AI agent governance that synthesizes diverse stakeholder perspectives
-Formation of an informal working group to continue multistakeholder dialogue on AI agent governance beyond the IGF
-Identification of key research gaps and policy questions that require further exploration
-Documentation of diverse regional and stakeholder perspectives on approaches to AI agent governance
Hybrid Format: Our hybrid engagement strategy ensures seamless integration of online and onsite participants:
We will employ a digital collaboration platform (Miro) where all participants can contribute ideas for AI agent governance principles in real-time, ensuring online voices have equal visibility.
Our dedicated online moderator will actively curate questions from remote participants, integrating them throughout the discussion rather than relegating them to the end.
We'll use live polling (Mentimeter) at strategic points to gauge alignment on emerging governance principles, with results instantly visible to all participants
Brief pre-session materials will be distributed to all registered participants to establish common understanding of key concepts.
The physical roundtable will include screens displaying online participants, with strategically placed microphones to capture in-room contributions.
For breakout discussions, we'll create hybrid groups that include both online and onsite participants, facilitated by group captains responsible for ensuring balanced participation.