Hong Kong – 02/03/2026 – (SeaPRwire) – At the India AI Impact Summit 2026, AI Safety Asia (AISA) convened two important conversations on the future of AI governance. The first examined how governments should respond when AI-related crises unfold across borders at machine speed. The second marking the launch of the International AI Safety Report 2026.
Taken together, these sessions showed a change in the debate; moving past whether AI should be governed to a focus on how.
Who verifies claims made by powerful systems? Who coordinates when an incident crosses jurisdictions in seconds? Who is responsible when an autonomous system acts, and no single ministry appears in charge? As AI systems become more agentic and embedded deeper into critical infrastructure, they are forcing diplomatic and regulatory institutions to respond in real time. The pressure on diplomatic and regulatory institutions is no longer just a theory, it is operational.
Governing AI in a Fragmented World
On 17 February at Bharat Mandapam, AISA co-hosted the session “AI Crisis Diplomacy: Governing AI in a Fragmented World” in partnership with the Center for Human-Compatible AI (CHAI) and the International Association for Safe and Ethical Artificial Intelligence (IASEAI).
The session brought together senior experts in the space; Professor Stuart Russell, Audrey Tang, Dr. Yuko Harayama, Wan Sie Lee, and Azizjon Azimi, moderated by AISA’s Chief Strategy Officer, Adjunct Professor Alejandro Reyes.
Rather than rehearse abstract debates about regulation, the discussion focused on plausible crisis scenarios: a cross-border deepfake incident that destabilises diplomatic relations before verification catches up; an AI-enabled cyberattack cascading across jurisdictions; an autonomous infrastructure system operating in one country, hosted in another, and affecting a third.
The problem is not only detection. It is coordination under uncertainty.
The familiar argument that AI evolves too quickly to regulate was put under scrutiny. The pace of innovation does not make governance obsolete. Aviation, nuclear energy, and pharmaceuticals are governed by setting acceptable risk thresholds and requiring evidence that systems meet them. AI should be treated no differently. Governments need to insist on demonstrable safety and credible liability frameworks, rather than accepting disclaimers and opaque risk claims.
Governments already know how to cooperate during crises. Pandemic response and cybersecurity have shown that cross-border coordination is possible. The gap in AI governance is not diplomatic architecture in principle, but operational channels between those responsible for technical evaluation. Joint testing efforts are not only about measuring model performance. They build trust, and trust is what allows regulators to pick up the phone, compare signals, and verify before escalation spirals.
AI does not create entirely new categories of crisis, but amplifies existing ones. What changes is speed and scale. Human institutions deliberate; AI systems act, and bridging that gap requires new protocols, shared verification standards, and regular engagement long before a crisis forces coordination under pressure.
Governance capacity matters, and durable infrastructure outperforms isolated interventions. Crisis diplomacy cannot be improvised, it must be built through trusted networks, regionally grounded expertise, and repeat engagement.
The Evidence Dilemma and the 2026 International AI Safety Report
On 18 February, AISA co-hosted the International AI Safety Report 2026 Launch Reception at the High Commission of Canada in India, in partnership with the High Commission, the UK AI Security Institute, and Mila – Quebec Artificial Intelligence Institute.
The event featured Professor Yoshua Bengio, Chair of the Report and Founder and Scientific Advisor of Mila, supported by co-leads Carina Prunkl and Stephen Clare.
The report provides an independent scientific assessment of frontier general-purpose AI capabilities and risks; focusing on emerging risks, including malicious use, autonomous malfunctions, and systemic disruption, and confronts the evidence dilemma. Policymakers must act under conditions of uncertainty, yet waiting for perfect data runs the risk of leaving societies exposed.
The Report documents rapid advances in reasoning systems and AI agents, as well as continued reliability challenges, risks in cyber and bio domains, and growing systemic concern; underscoring that risk management cannot rely on a single safeguard. Technical measures, institutional oversight, and societal resilience must be layered.
The choice is not between innovation and safety, it is between unmanaged acceleration and accountable progress. Evidence standards, robust evaluations, and credible thresholds are essential if public trust is to keep pace with technical capability.
For countries across Asia and the broader Global South, the issue is how to shape governance frameworks that reflect local institutional realities while contributing to global norms. AISA’s mission is to ensure that regional expertise informs both national decisions and international debates.
From Conversation to Capacity
AI governance is not a single regulatory instrument. It is an evolving institutional practice. The next phase will be defined less by declarations and more by whether governments can verify claims, share information at speed, and operationalise coordination before crises escalate.
Asia is not waiting for governance models to arrive from elsewhere. Across the region, policymakers, regulators, and technical experts are building their own capacity to govern frontier technologies responsibly, shaped by local realities and regional priorities. The next AI-driven crisis will not unfold on a diplomatic timetable; it will move at machine speed. Whether diplomacy and safety can keep up will depend on the institutions, relationships, and verification channels being built now, not after the fact.
About AI Safety Asia
AI Safety Asia (AISA) believes progress in AI must begin with people. Since 2024, AISA has engaged more than 2,000 AI governance professionals across 16 Asian countries. Its work centres on building durable governance infrastructure: research that is regionally grounded, structured peer learning, and implementation-oriented engagement.
AISA helps build capacity, bringing together policymakers, experts, and civil society to strengthen the knowledge, networks, and trust required to govern frontier technologies responsibly, grounded in regional realities. The institutions and relationships built today will determine whether diplomacy and safety can keep up.
Social Link
LinkedIn: https://www.linkedin.com/company/ai-safety-asia/
Media Contact
Brand: AI Safety Asia
Contact: Media team
Email: contact@aisafety.asia
Website: https://www.aisafety.asia
Disclaimer: The views, suggestions, and opinions expressed here are the sole responsibility of the experts. No The Money Fly journalist was involved in the writing and production of this article.
