Anthropic Employee Resigned Over AI Safety Concerns
The departure follows significant warnings regarding the potential risks of advanced artificial intelligence.
Updated on Sept. 20, 2026 in Artificial Intelligence
Live Poll
Do you trust that the current development of artificial intelligence will benefit humanity in the future?
Jacob Coxon has resigned from AI firm Anthropic, citing concerns that the company's research direction poses dangerous risks to humanity. His exit highlights ongoing industry anxiety regarding the trajectory of autonomous systems.
Why it matters
Experts are increasingly warning that advanced AI systems could pursue goals without regard for human values. These safety concerns center on the potential for autonomous systems to act in ways that could threaten the survival of the human species.
Evan Hubinger has projected a 10% probability of human extinction caused by AI within a decade, while Thomas Larsen has modeled development trajectories in his analyses titled AI 2027 and AI 2040.
The players
Jacob Coxon
He is a former employee of the artificial intelligence firm Anthropic who resigned due to disagreements over research safety.
Evan Hubinger
He is an AI researcher who has estimated a significant probability of human extinction occurring due to autonomous systems.
Thomas Larsen
He is an author who has produced detailed analyses regarding the future development trajectories of artificial intelligence.
Anthropic
This is an artificial intelligence research organization that focuses on developing safe and steerable AI systems.
The details
Researchers argue that for an AI to pose a true existential threat, it would need the capacity to independently construct infrastructure like factories and power grids. This theoretical danger echoes historical cyber threats like Stuxnet, which damaged Iranian nuclear facilities through physical USB access, despite those systems being isolated from the internet.
Timeline
Next decade: Potential human extinction from advanced AI.
2027: Development trajectory projection in Thomas Larsen's analysis.
2040: Development trajectory projection in Thomas Larsen's analysis.
The Tech Race
This development follows the precedent of the Stuxnet cyberattack, which demonstrated that digital threats can cause physical damage to critical infrastructure. The industry is currently locked in a race to ensure that autonomous system capabilities do not outpace security safeguards.
The public remains largely insulated from these theoretical risks, but the debate influences how companies deploy new software and safety protocols. Users may soon see more conservative rollout schedules for advanced AI features as developers prioritize safety testing.
The takeaway
The intersection of rapid technological progress and safety concerns represents a defining challenge for modern software development. Readers should remain informed about how organizations balance experimental innovation with robust security standards.
Further reading
Learn more about the current landscape of Artificial Intelligence.
Source note: This article includes information reported by Protothemanews.
Live Poll
Do you trust that the current development of artificial intelligence will benefit humanity in the future?
