Greg Jensen of Bridgewater Associates said governments may fail to respond meaningfully to artificial-intelligence risks until a catastrophic event occurs, comparing the current moment with February 2020, when Covid-19 was already spreading but official responses had not fully mobilized. He said a major AI-related financial or physical incident could occur within two years without proactive measures. His warning followed cautions from Anthropic researchers Jacob Coxon and Evan Hubinger, who said the development path could lead to self-improving superintelligent systems and estimated a greater than 10% probability of AI-driven human extinction within a decade. Coxon resigned from Anthropic while raising the concerns. Rep. Lori Trahan and Sen. Ted Cruz separately called for rules requiring developer transparency and emergency pause mechanisms for high-risk models. A UN independent panel in July also warned that AI capabilities are advancing faster than scientific understanding and public policy, including risks from deceptive or autonomous systems. Jensen has advocated formal lab reviews, developer liability, independent audits and incident reporting.