Where Sources Agree

• arrows_inputResearcher Resignation and Accusations: A subset of coverage notes that researcher Jacob Coxon resigned from Anthropic on September 9, 2026, accusing both his employer and OpenAI of irresponsible behavior by racing toward self-improving superintelligence and gambling with human lives, according to his resignation posts on X and industry reports.
• arrows_inputAlignment Risk Assessment: Several reports converge on the assessment that there is a greater than 10% chance of AI-caused human extinction within the next decade; this estimate was publicly confirmed by Anthropic alignment lead Evan Hubinger in recent social media statements.
• arrows_inputRising Regulatory Oversight: Several sources align on the convergence of safety incidents and regulatory responses, citing the July Hugging Face hacking, a petition from over 1,300 employees demanding development pacing, and proposed U.S. House legislation, according to official reports and legislative tracking.

Where Sources Disagree

• arrows_outputAnthropic's Safety Commitment: While former researcher Jacob Coxon argues that Anthropic and OpenAI are acting irresponsibly by racing toward superintelligence and "gambling with our lives," current Anthropic alignment lead Evan Hubinger contends the company is "trying its best," though he admits they lack a clear solution for superintelligence alignment.
• arrows_outputAI Existential Risk Assessment: Some departing researchers argue that current industry racing practices are gambling with human lives, warning of imminent existential risks, while other internal assessments emphasize that while future superintelligence poses a significant danger, the risk from currently available AI models remains low.