Anthropic Researchers Warn AI Could Cause Human Extinction by 2036
- Artificial intelligence researchers are warning that advanced systems could pose catastrophic risks to humanity, with one departing industry insider estimating a significant chance of human extinction by the...
- The debate over existential risk intensified after Jacob Coxon, an AI researcher at Anthropic, announced his resignation on social media, as reported by NewsNation.
- Coxon's remarks drew public support from colleagues within Anthropic.
Artificial intelligence researchers are warning that advanced systems could pose catastrophic risks to humanity, with one departing industry insider estimating a significant chance of human extinction by the end of the decade, according to public statements reported by NewsNation. The warnings underscore growing anxiety within the tech sector about the rapid acceleration of AI development. As companies build increasingly powerful models, developers are grappling with whether they can maintain control over future generations of the technology.
Inside the Warning From Resigning Researchers
The debate over existential risk intensified after Jacob Coxon, an AI researcher at Anthropic, announced his resignation on social media, as reported by NewsNation. Coxon stated that individuals building AI models earnestly believe the technology could kill off humanity by the end of the decade.
The people building AI earnestly believe it could kill us all by the end of the decade,
Coxon wrote on social media, adding that this is not a marketing stunt.
NewsNation
Coxon asserted that corporate executives and senior researchers often use more measured language in public while privately voicing grave concerns about safety. He maintained that no other human activity carries an equivalent level of danger.
Backing From Senior Alignment Leads
Coxon’s remarks drew public support from colleagues within Anthropic. Evan Hubinger, the company’s alignment-science lead, backed the assessment on social media.
Jacob is correct here — we really do earnestly believe AI could kill all humans!
Hubinger wrote in response, according to NewsNation. Hubinger added that he personally estimates the probability at greater than 10% within the next decade. He noted that while Anthropic is trying its best, the industry lacks a clear plan or trajectory to solve alignment for superintelligence.
Samuel Marks, Anthropic’s scalable-oversight lead, also weighed in on the discussion. Marks stated that developers in the field believe the technology could trigger human extinction or similar catastrophic outcomes within the next few years, noting that concern tends to increase alongside an employee’s seniority level, as reported by NewsNation.
Regulatory Pressures and Industry Incentives
The debate arrives amid heightened scrutiny over the pace of AI innovation and the adequacy of safety guardrails. Critics argue that major developers such as Anthropic and OpenAI might have financial motives to emphasize the dangers of their own technology, potentially using predicted risks to lobby for regulations that favor established firms over smaller competitors, according to NewsNation reporting. However, the researchers sounding the alarm work directly with these complex systems, including unreleased models that are not yet available to the public. That direct access has fueled demands from observers for policymakers to take the warnings seriously, even though precise risks remain difficult to measure or verify independently, NewsNation reported.