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

This is how Jacob Coxon said goodbye to the AI industry, on X, last week.

And in a follow-up post:

“The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.”

Jacob Coxon is not a star. He is not a CEO, not a famous scientist, and does not hold a PhD (yes, in USA you can be an “AI researcher” without a PhD). He joined Anthropic at 27 years old, mainly because of its safety reputation, and he likes to solve problems and “make a world a better place”.

He is, in other words, an engineer. Like most of us.

But he did something most of us never do: he left two months before his equity vested, and gave it up to speak.

“I no longer have anything to gain by juicing up Anthropic’s valuation. … I left before any of my equity vested.”

Note that he does not call Anthropic “evil”. He even told the Wall Street Journal that their safety work is sincere. His accusation is colder than that: “if you’re under pressure to race, you have to cut corners”.

You could argue the reaction was oversized. One resignation letter, more than 115 million views, a rage on social networks between engineers, entrepreneurs, or simply humans, and lawmakers suddenly calling for regulation.
But that is exactly the point: when the departure of a single engineer alarms the whole world, it tells you what the world already suspects: when the people inside the machine say this one is dangerous, then the noise is more a symptom than a bad buzz.

The insiders confirmed it themselves. Evan Hubinger, an alignment team lead at Anthropic, wrote:

“Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”

We have seen this movie before #

Five years ago, another employee looked at her own company’s internal research and faced the same choice.

Facebook’s own slides, revealed by the Wall Street Journal’s “Facebook Files”:

“We make body image issues worse for one in three teen girls.”

and:

“Among teens who reported suicidal thoughts, 13 percent of British users and 6 percent of American users traced the desire to kill themselves to Instagram.”

Frances Haugen did not shrug. She gathered the documents, resigned, and testified before the US Senate.

Coxon in 2026, Haugen in 2021. Different companies but same move: they refused, they left, and they warned the outside.

The pattern is the point. In both cases the company knew.
And, in both cases, someone inside had to decide whether knowing was enough.

The disease #

It is tempting to reduce this to a few evil companies. It is more uncomfortable, and more accurate, to look at the culture feeding them: ours.

Engineers today are asked to worship productivity and to treat quality as a luxury. A large 2026 survey of tech workers summarizes the situation in one respondent’s sentence:

“I can do more, faster, but not better.”

The same survey names the condition: smiling exhaustion.
People are shipping again, compensation has climbed, and there is no off-switch. More than half of tech workers (53%) would steer a newcomer away from their own field, and they stay. Another report on software delivery found that 67% of developers now spend more time debugging AI-generated code. We accept the slop, we accept the exhaustion, we accept watching the quality of our own work slip, all of that for the compensation.

This is not limited to AI labs. It is the same reflex that makes engineers rationalize surveillance products, addictive algorithms, defense contracts, and agency work with a shrug: “if I refuse, someone worse will do it”.

I wrote before that comfort is the enemy of skill, but I was too gentle because, at the end, we can see that comfort is the enemy of responsibility.

Trading morale for money is not a career decision but a transaction.
It is a transaction where you sell the part of you that was supposed to care. Do that often enough, and it does not come back.
That is not burnout, but it is still a disease. Although, like every disease, it spreads: first to your team, then to your product, then to the people using it.

The accusation #

So, why are you still working for them?

If the answer is the salary, say it louder, and do not force yourself to make excuses.
If the answer is “someone worse would replace me”, say it too, and louder too.
If the answer is “I can change things from the inside”, then show me what you changed. Haugen tried the inside. Coxon tried the inside. At the end, unfortunately, the inside did not move.

We live in a society, and every act we take carries a responsibility.
You can ignore it, collect your compensation, keep your head down, and tell yourself you are just an engineer building features… But then, do not consider yourself a contributor to humankind. You are just a human thinking about itself, nothing more.

The people who refuse are not heroes but human, and good people refuse to do bad things.