// KENDDRICK CHAN Kenddrick Chan is head of the Technology and International Affairs project
For policymakers, there is a qualitative difference between governing genera- tive and agentic AI. Agentic AI is akin to deploying thousands of self-driving vehicles on roads with no shared traf- fic rules, with each vehicle optimised for different goals (such as speed, safety, effi- ciency or cargo integrity). The risk posed is much less of individual system fail- ure (that is, any single vehicle is faulty) but much more of the systemic instabil- ity that emerges from the interaction of such systems at scale (in other words, the uncoordinated and unpredictable char- acter of interaction that increases the risks of vehicle accidents). The relevance of agentic AI is further amplified by the fact that it is increas- ingly discussed as a potential engineering pathway towards artificial general intelli- gence, insofar as it introduces capabilities such as goal decomposition, tool use, iterative self-correction and multi- step execution. While this does not yet amount to a proven route to artificial gen- eral intelligence, it nonetheless reinforces the strategic significance of agentic sys- tems. If managed well, however, the same properties that make agentic AI systems potentially destabilising could become sources of great collective benefit. A well-governed ecosystem of AI agents that operate under shared governance pro- tocols, with clear accountability chains and interoperable safety standards, could deliver upon AI’s development poten- tial. The question for G7 governments is whether they can agree upon the urgently needed rules of the road. FROM NATIONAL FRAMEWORKS TO TRANSNATIONAL RISKS Although G7 leaders collectively reaf- Roadmap at their Kananaskis Summit in 2025, individual G7 governments have moved faster. The United States published its AI Action Plan, the UK advanced its AI Security Institute’s evaluation work, and the European Union’s AI Act began impos- ing binding obligations on providers of general-purpose AI models. Yet rules for- mulated at the national level, however robust and well intentioned, cannot suf- ficiently match a technology whose most consequential deployments are by nature transnational. A multi-agent system, whether coordinating a disinformation firmed the Hiroshima AI Process commitments via its AI Adoption
campaign or managing financial port- folios, does not pause at the border of national jurisdictions.
SEIZING THE WINDOW FOR COORDINATED GOVERNANCE
With regard to agentic AI, two challenges remain critically urgent yet unresolved. First, there are no agreed-upon inter- national standards for evaluating the pre-deployment safety of agentic sys- tems, particularly for high-risk sectors such as critical infrastructure, defence and health care. Second, liability frame- works have not kept pace, as the question of who bears legal responsibility when an autonomous agent causes harm – whether it is the developer, deployer or operator – remains deeply contested. At the Évian Summit, G7 leaders should make progress on providing the architecture needed for collectively gov- erning agentic AI. Two deliverables would constitute decisive progress. The first is a commitment to accept each other’s pre-deployment assessments for high-risk agentic AI applications, par- ticularly in areas such as autonomous cyber defence or military decision- support systems. Such mutual recog- nition regarding agentic AI safety evalu- ations would be a good first step towards mitigating the critical dangers of regula- tory fragmentation in the absence of full harmonisation, dangers such as failures in infrastructure management or unin- tended security escalations. The second deliverable is a shared reporting mech- anism, specifically regarding agentic AI incidents in critical sectors. This would serve as the foundation for a joint G7 early warning system that could provide much-needed situational awareness. The window for meaningful coop- eration is closing. As proprietary dependencies harden, the eventual imposition of common standards will likely be more disruptive and thus more fiercely resisted. Given that subsequent regulation will only become more chal- lenging, the Évian Summit offers G7 leaders the chance to show that dem- ocratic governments can match the pace of the technology they are trying to govern. This matters not only for AI safety but also for the credibility of the G7 itself as a group capable of managing the defining technological challenges of the times ahead.
at LSE IDEAS, the foreign policy think tank of the London School of Economics. He is also a doc- toral candidate at the University of Oxford and was a 2022–2023 Fellow at the Portulans Institute.
// CHRIS ALDEN
Chris Alden is a professor of international relations at the London School of Econom-
ics and Political Science and the director of LSE IDEAS. He was a co-chair of the G20’s Task Force on Reformed Multilateralism for India’s G20 2023 presidency.
// CORNELIU BJOLA Corneliu Bjola is a professor of
digital diplo- macy at the University of Oxford and the head of the Oxford Digi- tal Diplomacy Research Group. His current research examines the emerging role of artificial intelligence and quantum tech- nologies in diplomacy, with a focus on their capacity to sup- port decision-making, refine negotiation practices and strengthen crisis response, while also addressing the ethical and governance challenges these technologies pose across diplomatic institutions, processes and international order.
119 globalgovernancemedia.org
Made with FlippingBook - professional solution for displaying marketing and sales documents online