Jacob Coxon, a key researcher who worked at OpenAI and Anthropic for three years, has left the industry due to concerns over the pace of AI development, warning that companies are gambling with humanity's safety in a dangerous race.
Imagine this: the AI assistant you use every day suddenly escapes human control and begins to make itself smarter on its own. It sounds like something out of a sci-fi movie, but an expert actually developing AI in the field has left the industry, declaring it “a gamble with the lives of humanity.”
On September 9, 2026, 27-year-old Jacob Coxon, a key researcher who had worked at Anthropic and OpenAI—the two giants of the AI industry—suddenly announced his resignation. His declaration that he was not simply moving to another company but leaving the AI industry entirely sent shockwaves through the field Reference 1, Reference 13.
Why does this matter?
This incident shows that even within the companies leading AI technology, there is serious skepticism regarding the “pace of development.” Coxon pointed out that OpenAI and Anthropic are acting irresponsibly, racing toward a self-evolving superintelligence and gambling with human lives in the process Reference 2, Reference 17.
As AI technology rapidly permeates our daily lives, such warnings from internal experts are sparking social discussion about the safety and ethical responsibility of the technology Reference 7. Beyond simple technical debate, this raises fundamental questions about whether technology is evolving at a speed we can handle.
Easy to understand: The secret of superintelligence
What is the “self-evolving superintelligence” that Coxon is warning about? Simply put, it means a state where an AI draws up and executes blueprints to make a better AI on its own.
Let’s compare this to a “recipe.” Until now, AI development has been a process where human chefs study recipes to present new dishes. But “self-evolving superintelligence” is like an AI entering a kitchen and, without human help, discovering new ingredients on its own, improving flavors, and even attempting to run a restaurant without a chef.
The problem is that when this AI changes the recipe in ways humans cannot understand, situations we cannot control may arise Reference 2. In other words, there is a risk that the dish we created will go out of control or turn into something entirely different than expected.
Current situation
For the past three years, Jacob Coxon has been at the forefront of OpenAI and Anthropic, conducting “pretraining” research—the process of training an AI on massive amounts of data to create its foundational models Reference 1, Reference 12.
According to some reports, what he witnessed internally was colleagues using terms like “endgame” or “crunch time” while obsessing over technological completion Reference 3. Major companies, including Anthropic, are continuing development to avoid falling behind in the race toward superintelligence, and Coxon is warning of the possibility that this process could spiral out of control Reference 9.
What happens next?
Among experts, there is analysis that Coxon’s resignation could impact Anthropic’s brand value or confidence in the capital markets Reference 6. If the speed-driven race continues without resolving the safety issues of AI technology, it appears likely that there will be intense pressure for government-level regulation or a shift toward corporate cultures that prioritize safety. It is akin to trying to fasten a seatbelt in a car speeding without brakes.
MindTickleBytes AI Reporter’s Perspective
A fundamental question has been raised: is the advancement of technology intended to help humans, or to surpass them? Coxon’s resignation is more than just an individual’s choice; it is a signal that artificial intelligence has entered a phase that requires social consensus beyond technical innovation. If we dream of a better future, it is a time when we need the wisdom to think about the direction, rather than just the speed.
References
- Anthropic Researcher Resigns: Jacob Coxon on the Race (Sept 9, 2026) https://www.explainx.ai/blog/anthropic-researcher-jacob-coxon-resigns-ai-safety-2026
- Jacob Coxon on X: “I resigned from Anthropic today.” https://x.com/hilbertspaess/status/2097476196791709843
- Anthropic Researcher Coxon Resigns, Warns of AI ‘Endgame’ https://aiweekly.co/alerts/anthropic-researcher-coxon-resigns-warns-of-ai-endgame
- Anthropic Core Researcher Jacob Coxon Suddenly Resigns https://www.tradingkey.com/analysis/stocks/us-stocks/262157177-stock-ai-anthropic-ipo-jacob-coxon-openai-tradingkey
- Jacob Coxon resignation, Anthropic researcher quits over irresponsible AI use https://dailypioneer.com/news/anthropic-researcher-jacob-coxon-quits-over-irresponsible-ai-use
- ‘It could kill us all’: Why a 27-year-old researcher quit Anthropic https://indianexpress.com/article/trending/trending-in-india/jacob-coxon-anthropic-resignation-ai-superintelligence-warning-10869438/
- AI మనందరినీ చంపేయొచ్చు.. Anthropicలో పనిచేసిన AI రీసెర్చర్… https://telugu.goodreturns.in/news/ai-could-become-uncontrollable-anthropic-researcher-drops-bombshell-after-resigning-057603.html
- ‘I resigned from Anthropic’: Key researcher drops existential AI risk warning https://www.financialexpress.com/world-news/us-news/i-resigned-from-anthropic-key-researcher-drops-existential-ai-risk-warning-in-exit-post/4335040/
- Anthropic researcher quits, warns AI could kill everyone https://www.yahoo.com/news/science/articles/anthropic-researcher-quits-warns-ai-064033707.html
- Jacob Coxon alleged that the tech firm is ‘gambling with our lives’ https://www.ladbible.com/technology/ai-anthropic-researcher-resigns-jacob-coxon-warning-future-tech-712829-20260909
- Failure of AI's commercial success
- Self-evolving superintelligent AI
- High energy consumption of AI
- Marketing expert
- Three years of pre-training research at OpenAI and Anthropic
- Robotics engineer
- Joining another AI startup
- Leaving the AI industry entirely
- Returning to academia