AI risk discourse intensifies as experts warn of existential threats

Former Anthropic researcher Jacob Coxon has raised alarms about artificial intelligence posing an extinction-level risk to humanity in the coming years. Coxon, who spent three years researching at Anthropic and OpenAI, has been promoting his warnings since his resignation, emphasizing that an intelligence explosion—where AI systems create more advanced versions that could surpass human intelligence—could lead to catastrophic outcomes, including infrastructure sabotage or bioweapon development. The International Monetary Fund has convened tech and finance experts to explore how AI-related disruptions could trigger global economic instability. Amid these discussions, a growing lexicon of AI risk terminology is emerging, aimed at clarifying the complex landscape of potential threats. Evan Hubinger, Anthropic’s alignment science lead, stated there is a greater than 10% chance of an existential risk occurring within the next decade.