Anthropic 2026 Exit: AI Could Kill Us, Crypto AI Responds

AI researcher Jacob Coxon resigned from Anthropic, warning frontier labs are gambling with humanity by racing toward self-improving superintelligence despite private existential fears, a shift intensifying crypto's push for decentralized, verifiable AI infrastructure and governance amid growing regulatory scrutiny.

After spending three years at OpenAI and Anthropic, AI researcher Jacob Coxon has left Anthropic with a very strong warning that, in fact, many top executives in private conversations already believe that superintelligence may end in an existential disaster as early as the end of this decade if it goes uncontrollably. Still, they race ahead to create systems that will be able to improve themselves automatically.

Labs Risking Humanity

He said that the so-called frontier labs were risking human safety as they were playing with highly developed machines that were capable even of making changes by themselves. At the same time, their company staff was raising the issue internally. His decision to walk away comes in the wake of warnings issued by former employees of both OpenAI and DeepMind.

Source: Reuters

Also Read: Anthropic IPO Nears as $65 Billion Revenue Signals Strong AI Demand

Why Crypto and Decentralized AI Should Pay Attention

Blockchain is under pressure from the shift of power away from a central system over computational resources, big data, and financial infrastructure, which goes beyond the issue of ethics to the stability of the system.

A centralized superintelligence could undermine the very idea of decentralization as practiced by crypto networks. This is particularly so as developers building the decentralized AI networks, which include Bittensor, Render, Fetch.ai, etc., are grappling with major challenges in alignment, verification, and censorship resistance.

I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
— Jacob Coxon (@hilbertspaess) September 9, 2026

So, for every stakeholder of these new technologies – whether they be the creators of the decentralized AI systems, the purchasers of AI tokens, the platforms listing these new financial instruments, or the DAO designers who govern the rules – there is a lot more at stake than just whether the technology will work.

Also Read: OpenAI vs Apple Lawsuit 2026: Trade Secrets Battle Heats

AI Crypto Faces Reckoning

At a time when AI-related crypto assets already top $30B on CoinMarketCap, Anthropic’s warning comes in time. The surge in venture financing for decentralized compute can be seen as part of the bigger pushback against centralized labs like Anthropic.