An Anthropic researcher named Coxon resigned, saying on X that the people building AI believe it could kill everyone by the end of the decade. The warning, compared to The Terminator, highlights growing fear inside AI labs.
What actually happened
According to CoinDesk, a researcher identified as Coxon resigned from Anthropic and explained the decision on the social platform X. 'The people building AI earnestly believe that it could kill us all by the end of the decade,' Coxon said. The comment drew comparisons to The Terminator, a science fiction film in which machines turn against their creators. The report does not specify Coxon's job title, tenure at Anthropic, or the exact date of resignation. Anthropic has not issued a public comment addressing the claim, based on the material available. The headline framing, which echoes the script of The Terminator, was used by CoinDesk to describe the tone of Coxon's warning, not a direct quote from Coxon himself.
How we got here
This resignation adds to a broader conversation about the pace of AI development and the confidence of insiders in the technology they help build. The claim itself, that AI could threaten human survival within years, is presented as a belief held by unnamed people building AI, not as evidence or data. The report offers no technical basis for the timeline given. It also does not identify how many people share this view inside Anthropic or elsewhere. The comparison to The Terminator underscores how public discourse around AI risk increasingly borrows from science fiction rather than technical papers.
Why this matters for you
For builders, this resignation is a reminder that AI safety concerns exist even inside leading labs, not just among outside critics. For everyday users of AI tools, it raises the question of how much trust to place in systems built by teams who may privately doubt their own safety timelines. For crypto and Web3 builders exploring AI integration, including AR and smart glasses hardware that increasingly relies on AI models, this is a signal to watch how safety debates evolve before deeper integration. For investors, sentiment around AI stocks and AI-linked tokens could shift if more insiders speak out publicly.
The bigger question
If people inside the companies building AI privately believe it could threaten humanity within years, what obligation do those companies have to slow down? Who decides when the risk of continued development outweighs its potential reward, and should that decision rest with company insiders, governments, or the public whose lives may ultimately be affected by the outcome of this technology?
What to watch
No official date has been set for further comment from Anthropic. Coxon's own timeline, 'by the end of the decade,' points to around 2030 as an informal marker for AI risk debates. Bonuz will watch how AI safety statements from insiders at major labs evolve, especially as AI increasingly powers wearable and AR devices entering the market.
