Sam Altman may be at the top of the world of artificial intelligence, but former employees claim he doesn’t fully understand the technology he’s selling. A recent investigation uncovered surprising claims about OpenAI CEO. The man who runs the largest AI company on the planet often gets the basic concepts of machine learning confused, according to people who have worked with him.
These revelations portray a leader who operates primarily as a businessman rather than a technology innovator. Insiders claim that Mr. Altman doesn’t fully understand the underlying mechanics of the software he sells, a disconnect that raises questions about the company’s direction.
Former employee talks about Altman’s technical knowledge
CEOs are far from technical experts. Those around him say he doesn’t have deep knowledge of coding or building machine learning models. This gap is becoming increasingly apparent to engineering teams building enterprise systems.
Multiple developers recalled specific instances where Mr. Altman misused or confused basic technical terminology. Non-technical CEOs are not uncommon in Silicon Valley. Still, OpenAI’s size is so large, currently valued at around £63.5 billion ($80 billion), that the knowledge gap is a concern for investors who place a lot of trust in leaders who rely on their staff to build the product.
Convince people instead of writing code
What sets Altman apart is his talent for persuading engineers, investors, and the public to support his vision. He regularly convinces these disparate groups that their different priorities are his priorities, and this diplomatic skill has proven more valuable to the company than any programming language.
Whenever his internal critics try to challenge his next move, he always finds words to neutralize them. His approach is to set boundaries to appease the parties involved, then tear them down later.
“He’s creating structures that, in theory, will constrain him in the future,” said former OpenAI researcher Carol Wainwright.
A recent exposé in The New Yorker claims that OpenAI CEO Sam Altman has no idea how his company’s AI technology works.
Former OpenAI employees argued that he didn’t really understand basic machine learning concepts and didn’t fully understand the products he was selling to people. pic.twitter.com/ShRgFsnXb3
— Publicity (@pubity) April 11, 2026
Insiders talk about Altman’s approach to safety structures
These structures are designed to reassure safety-conscious staff about responsible development. However, it appears conditional when it conflicts with expansion goals.
Mr. Wainwright detailed how CEOs deal with self-imposed limitations once technology advances. “But when the future comes and he is constrained, he removes all previous structures.”
These findings come from an 18-month investigation by Ronan Farrow and Andrew Marantz for The New Yorker, based on more than 100 interviews and previously unpublished internal documents. IBTimes UK separately reported on the findings of an extensive investigation into Altman’s conduct at OpenAI.
On April 10, Altman published a blog post describing the New Yorker article as “inflammatory” and said it “underestimated the power of words and stories.” He admitted that he has a tendency to be “conflict averse,” which he said has “caused me and OpenAI a lot of pain.” His post came after someone threw a petrol bomb at his San Francisco home earlier in the day, an incident that police linked to a suspect who was later arrested at OpenAI’s headquarters. There were no injuries.
