- Three current and former employees at the American artificial intelligence firm Anthropic have publicly warned that advanced AI technology could lead to human extinction within the next decade.
- Evan Hubinger, Anthropic's alignment-science lead, stated on X that he estimates a greater than 10 per cent probability of AI killing humanity within ten years, warning that developers currently lack a plan to align superintelligence.
- Former researcher Jacob Coxon announced his resignation from Anthropic and asserted that senior industry figures privately express immense fear over existential risks while softening their public statements.
- Samuel Marks, a scalable-oversight lead at Anthropic, echoed these warnings in a personal capacity, noting that concern over catastrophic outcomes is generally higher among more senior AI developers.
- Anthropic has been contacted for comment, while rival developer OpenAI declined to comment directly on the statements, pointing instead to its recently published AI safety and alignment updates.
IN FULL
