Professor B Ravindran, head of the Centre for Responsible AI at IIT Madras, has been appointed as the only Indian member of the world's first global scientific body on AI. The 40-member Independent International Scientific Panel on AI will produce periodic scientific assessments to guide the United Nations Global Dialogue on AI, a forum where governments negotiate how to govern artificial intelligence.
The appointment, confirmed June 21, 2026, places an Indian voice at the center of a debate with high stakes for developing nations. Countries without the infrastructure to build AI systems risk becoming what Ravindran calls "digital colonies" if global rules are written without their interests in mind.
What the scientific panel actually does
The panel's mandate is narrow by design. "Our primary responsibility would be to produce reports on the scientific aspects of AI," Ravindran said. "This body would not get into the politics or policy side of AI discussions. That is the responsibility of the Global Dialogue, in which governments would participate."
The group takes a periodic view of the state of AI for Science & Research, delivering evidence that informs diplomatic negotiations. It is the first time a standing scientific body on AI has operated at this global scale.
The case for global rules, not fragmented ones
Ravindran sees two problems with the current country-by-country approach to AI regulation. First, fragmented rules slow development. Companies must engineer their systems to comply with conflicting requirements across jurisdictions, and they may choose to deploy services only in countries with friendlier regulatory climates.
Second, many nations lack the resources to write strong regulations at all. "Many countries in Asia or Africa might not have the resources or capacities to frame robust regulations to protect their interests," he said. "There would always be a danger of them being reduced to digital colonies."
A globally agreed minimum set of regulations, he argues, would protect countries that cannot build AI themselves while preserving their access to its benefits. "AI is a very transformative technology, with a transformative capability of the order of the steam engine. It would be awful if, owing to a lack of regulatory capacities, certain developing countries were denied the benefits of AI."
The risk of a non-proliferation regime for AI
Ravindran acknowledged the logic behind international controls on AI models that could enable biological or chemical weapons. But he warned that the same reasoning could produce a regime resembling nuclear non-proliferation treaties - where only a handful of countries or companies are trusted to develop advanced systems.
"I am worried about the dialogue on AI moving in that direction," he said. "Hopefully, it would not happen, but I think we have to be on constant vigil against it."
India's Trusted AI Commons
A separate piece of the governance puzzle emerged from the New Delhi AI Impact Summit in February. The Trusted AI Commons, hosted and managed by India, will serve as a repository of tools, benchmarks, datasets, and protocols for testing AI systems safely.
Ravindran described it as "a one-stop shop to figure out what tools are available to test AI deployment." The Commons will not commission new tools immediately, but will aggregate existing resources - including work from IIT Madras's Centre for Responsible AI and tools released by companies like Google. It is called a Commons because access will be open, with liberal licensing requirements.
The initiative aligns with broader discussions about how AI for Government and public-sector institutions can build shared infrastructure rather than relying entirely on proprietary systems controlled by a few firms.
Why this matters for science and research professionals
The creation of a standing international scientific panel on AI signals that AI research is now treated as a domain requiring the same kind of sustained, evidence-based global assessment as climate science. For researchers, the panel's reports will shape which questions get funded, which risks get prioritized, and which tools become accessible through shared infrastructure like the Trusted AI Commons. The direction of these discussions - toward open access or toward restricted, non-proliferation-style controls - will directly affect who can participate in AI research and under what conditions.
Your membership also unlocks:
