Anthropic researcher estimates over 10% risk AI could cause human extinction

A new advisory issued this week marks the latest in a growing series of warnings about the safety threat posed by artificial intelligence. The notice, released by the Future of Life Institute in collaboration with leading AI researchers, cautions that rapid advances in generative models and autonomous systems could outpace current oversight mechanisms, increasing the risk of unintended harmful outcomes. The institute’s statement follows earlier alerts from the European Commission, the U.S. National Security Commission on AI, and prominent technologists who have highlighted concerns over alignment failures, data privacy breaches, and the potential for AI-driven disinformation campaigns.

The advisory calls for immediate, coordinated action to strengthen governance frameworks, improve transparency in AI development, and invest in robust safety research. It urges policymakers to adopt clear standards for testing and verification before deployment, and recommends that industry leaders allocate a minimum of 10 percent of AI project budgets to risk assessment and mitigation. As the dialogue around AI regulation intensifies, the latest warning underscores the urgency of establishing safeguards that can keep pace with the technology’s accelerating capabilities.