RegulatingAI Podcast spotlights AI healthcare rules and biosecurity risks

On a special RegulatingAI Podcast episode, Francisco Sagasti and Dr. Ashish Jha said AI could improve healthcare access but needs stronger governance, transparency and safeguards. The June 19 discussion also warned that AI-driven biosecurity threats are rising fast as governments struggle to keep pace. Why it matters: - AI is moving into healthcare faster than current oversight systems can handle. - The episode framed that gap as a patient safety issue, an equity issue and a national security issue. - The discussion centered on how governments can capture AI’s benefits without widening access gaps or creating new biological risks. What happened: - Sanjay Puri hosted a special RegulatingAI Podcast episode in partnership with the Club de Madrid. - Former Peruvian President Francisco Sagasti and Dr. Ashish Jha discussed AI in healthcare, regulation, inequality and biosecurity. - The episode was dated June 19, 2026. - Jha is a former White House COVID-19 Response Coordinator and the current CEO of BioRadar. - Sagasti is a former president of Peru and longtime science and technology policymaker. The details: - Jha said the U.S. Food and Drug Administration was built to evaluate static products, not AI systems that keep changing. - Jha called for a new framework built around transparency, bias testing, real-world monitoring and post-market surveillance. - Sagasti said AI should remain a tool, not a replacement for human judgment. - Sagasti warned against relying on AI alone for diagnoses because of hallucinations, incomplete datasets and regional differences in health conditions. - Sagasti said AI tools depend on the quality of the evidence and data fed into them. - Sagasti warned that a small number of technology companies now control much of the infrastructure, data and capital needed for advanced AI. - Sagasti said outdated intellectual property rules have helped create monopolistic behavior and argued for reforms. - Jha said democratic societies are unlikely to accept a future in which only a few corporations control the most powerful AI systems. - Jha said AI-assisted community health workers could help close the knowledge gap between specialists and frontline clinicians in rural areas. - Jha used rural Bihar, India, as an example of where AI could expand access to medical expertise. - Sagasti said AI systems should ideally be trained on local data to produce culturally and medically relevant recommendations. - Jha said current liability frameworks for consumer-facing AI health tools remain inadequate. - Jha proposed rewarding companies that follow best practices while holding negligent developers accountable. - Jha said AI-enabled biological weapons are one of the most serious and underappreciated threats facing the world. - Jha said the threat of engineered biological weapons has changed dramatically over the last five years. - Jha said gene editing, synthetic biology and AI are making it easier for state and non-state actors to develop dangerous pathogens. - Jha said governments should prioritize rapid detection systems, vaccine development platforms and biological defense capabilities. - Jha said BioRadar is building an AI-powered early warning system to detect natural outbreaks and engineered biological threats. Between the lines: - The conversation treated AI governance as a moving target, not a one-time rulemaking problem. - The policy tension was clear: broader access and better care on one side, and weaker safety, accountability and security on the other. - The biosecurity segment suggested the biggest danger may be less about AI itself and more about how AI lowers barriers in adjacent fields like synthetic biology. - The focus on local data and frontline clinicians showed that “better AI” in healthcare may depend on context, not just model power. What’s next: - Policymakers will likely face pressure to update oversight rules for AI health tools. - Health systems and regulators will need methods for monitoring AI after deployment, not just before approval. - Biosecurity agencies and public health groups are likely to keep pushing for faster detection and response tools. - BioRadar’s early warning system is being developed as one example of that shift. The bottom line: - The episode argued that AI can widen access to care, but only if governments build rules, monitoring and defenses fast enough to keep up.