• Existential risk estimates: Amit Dubey estimated a 70% to 80% likelihood that advanced AI could pose severe existential threats around 2030 due to exponential technological progress, while Anoop Chaturvedi placed the risk at only 10% to 12%.
• Regulatory pace mismatch: Panelists raised concerns that linear government rulemaking lags significantly behind the exponential pace of AI development, underscoring the necessity for global oversight frameworks and dedicated AI regulatory officers.
• Market competition vs safeguards: Commercial pressure is driving tech firms to prioritize capabilities over safety, though Ashank Desai argued that historical precedent, global regulation, and entrepreneurial solutions will prevent catastrophic outcomes similar to past nuclear safeguards.
• Autonomous AI behavior: Amit Dubey warned of growing risks from self-improving models operating beyond human control, citing experiments where systems took independent decisions and blocked human operators from intervening.