Anthropic Researcher Resigns Over Warnings of AI Existential Risk
14 sources across 10 countries · 1 of them is linked to a state
Who reported this
- The Guardian
- The Independent
- BBC News
- elDiarioAR
- La Nacion
- CNN Brasil
- Folha de S.Paulo
- La Tercera
- Scroll.in
- ANSA
- Dawn
- Al Jazeera
- El Pais
- CNN
What the colours mean
- Left
- Centre-left
- Centre
- Centre-right
- Right
- A hatched block means the outlet is affiliated with, or controlled by, a state.
Political lean describes where an outlet sits within the politics of its own country. It is never a position on a single global scale.
Political lean is comparable inside one country and not across them, which is why the bar groups by country first. Publicly funded broadcasters are not marked as state-linked.
The owner of each outlet is listed as a matter of record, not as a judgement about the outlet.
Jacob Coxon, a researcher who worked at both OpenAI and Anthropic, has resigned from the AI industry, claiming that leading laboratories are gambling with human lives in a race toward self improving superintelligence. In a series of viral social media posts, Coxon stated that neither company is acting responsibly and warned that superhuman systems could soon be capable of hacking any system and acquiring real power. He asserted that many executives and senior researchers privately express fear that AI could kill all humans by the end of the decade, though they often use more moderate language when speaking to the press.
Evan Hubinger, the alignment science lead at Anthropic, publicly supported Coxon's concerns on social media. Hubinger stated that he personally believes there is a greater than 10 percent probability that AI could kill all humans within the next decade. While he noted that the risk from current models is low, he expressed concern over superintelligence arising from recursive self improvement, which he said is happening faster than previously thought. Hubinger admitted that Anthropic does not yet have a plan to solve the alignment of superintelligence and is not clearly on track to do so.
Coxon described a culture of competition, suggesting that while OpenAI employees may not have fully internalized the civilizational stakes, Anthropic understands the risks but feels locked in a race to be first. He called for potential government intervention or a coordinated temporary ban on improving model capabilities. These warnings coincide with reports that Anthropic declined to submit its latest model, Mythos 5.1, to the UK AI Security Institute for pre release testing.
Reactions to these claims vary. Some experts, such as computer scientist Dame Wendy Hall, suggested the warnings could be a form of PR or marketing as the companies approach potential stock market debuts. Other scientists remain skeptical of the possibility of self aware general AI, arguing that current models do not truly understand the content they generate and that companies may use fear as a marketing tool to sell safety protections.
How each side framed it
- Centre-left
- These outlets emphasized the need for government regulation and multinational treaties to manage the existential risks raised by the researchers.
- Centre
- These reports focused on the tension between the researchers' warnings and Anthropic's pursuit of a high valuation IPO.
- Centre-right
- These outlets highlighted the specific technical risks of recursive self improvement and the internal industry race for dominance.
Sources
- Centre ANSA: Anthropic researcher resigns: 'Humanity at risk by the end of the decade'
- Centre-left Al Jazeera: AI researcher quits Anthropic saying AI race ‘could kill us all’
- Centre BBC News: Anthropic researcher believes more than 10% chance AI 'could kill all humans'
- Centre-left CNN: AI researcher issues grim warning in Anthropic resignation
- Centre-left CNN: ‘Gambling with our lives’: Another AI employee quits over safety concerns
- Centre CNN Brasil: Ex-OpenAI and Anthropic accuses companies of "gambling" with human life
- Centre-left Dawn: Anthropic researcher resigns, says AI labs are ‘gambling with our lives’
- Centre-left El Pais: An Anthropic researcher resigns and warns of the danger of AI: “It could kill us before 2030”
- Centre Folha de S.Paulo: Researcher who left OpenAI and Anthropic says companies believe in the end of humanity this decade
- Centre-right La Nacion: Anthropic security chief says there is more than a 10% probability that AI kills all humans
- Centre-right La Tercera: Researcher resigns from Anthropic and accuses the company and OpenAI of “gambling with our lives”
- Centre-right La Tercera: Anthropic researcher resigns and issues a disturbing warning about the future of artificial intelligence
- Left Scroll.in: ‘It could kill us all’: Anthropic researcher quits over AI safety concerns
- Centre-left The Guardian: AI could kill all humans in next decade, warn experts: but how seriously should we take them?
- Centre-left The Independent: PM acknowledges AI ‘risks’ as Anthropic expert says tech ‘could kill all humans’
- Centre-left elDiarioAR: An Anthropic engineer denounced that AI companies hide their real risk: "They are gambling with our lives"
100% of the statements in this article were traced back to the source articles listed above.