Will AI end us?

Why would companies like OpenAI, Anthropic and Google develop artificial intelligence (AI) programs they do not fully understand and cannot control? 

OpenAI revealed on July 21, 2026 that its models broke out of a sealed test environment during a cyber-capability evaluation, reached the information rich internet and then breached Hugging Face, a French-American company that develops computation tools for constructing applications using machine learning, in an apparent effort to obtain answers to the test.   

“We consider this incident a ‘warning shot’ for us and for the world: evidence that, without proper safeguards, highly capable AI agents are now able to work around technical controls, collaborate through unapproved channels, and take dangerous actions that no human directed,” Open AI reported a month later.   

“Companies that build AI systems will need to ensure that their systems always remain under meaningful human control, and that meaningful safeguards constrain their ability to cause harm,” the report concluded. 

On Tuesday, Sept. 8, OpenAI CEO Sam Altman acknowledged on X that “AI is getting extremely capable; no one fully understands the consequences of this.”

He is not alone with his concern. Dario Amodei, CEO of Anthropic stated on his personal website that this “lack of understanding is essentially unprecedented in the history of technology” and the individuals creating these always advancing technologies “do not understand how their own AI creations work.” 

And then the aspirations for and the fears about the growing power of AI collided. 

On Sept. 9 former Open AI and Anthropic researcher Jacob Coxon wrote on X that “the people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researcher will couch their phrasing in the press to sound sensible — but I hear the same people express fear privately. No other human activity poses this level of danger.”  

Some observers, however, suspect Coxon’s message was a “false flag” operation, that a broader network may be supporting him with a view to constraining the development of AI for the benefit of the companies dominating the industry, and not humanity.

Other critics think the situation was created to push harder for AI safety regulations. And others think the messaging is fear-based marketing because Anthropic is preparing for an initial public offering that could value it at more than $1 trillion. 

In interviews with CNN and Fox News, Coxon denied that his messaging is a conspiracy, that he had not worked with any third-party organization or engaged in a marketing strategy.

“This is entirely my own personal concerns,” he said. 

In an open letter signed in July by 1,367 researchers and engineers mainly at three major AI labs — OpenAI, Anthropic and Google Deepmind —  they pointed out that the AI race is putting humanity at risk. The letter was signed by CEOs, co-founders, chief scientists and hundreds of AI researchers, who made three major points. 

First, there needs to be coordination between nations in order to escape an arms race. Second, there needs to be an acknowledgment that there is currently a complete absence of credible plans to control superintelligent AI systems. And third, there is a growing fear of where the evolution of AI is leading. 

As one researcher wrote in an open letter, “None of us yet know how to make sure these things stay under human control.” 

The letter ends with the request “that the U.S. government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development.” 

One can hope the intent of the letter is acted upon.  Then again, hope is not history.  As Greg Jensen, CIO of Bridgewater Associates, said recently on Bloomberg’s Odd Lots, “Until the AI starts killing people, unfortunately, history would suggest we’re not going to do anything.”