Bayoumi column links Altman's 'bad things' remark to reported AI deaths

White House ‘Super Intelligence’ accord signed by Trump, six tech CEOs

The Guardian published an opinion column on Thursday by Moustafa Bayoumi, a Guardian US columnist and professor of English at Brooklyn College, City University of New York, that ties OpenAI chief executive Sam Altman’s recent statement that society should accept “some bad things” from artificial intelligence to a series of reported harms the columnist attributes to the technology. Bayoumi is the author of the award-winning books “How Does It Feel to Be a Problem?: Being Young and Arab in America” and “This Muslim American Life: Dispatches from the War on Terror.”

Bayoumi opens by quoting Altman from Politico’s technology podcast and newsletter, Decoded: “One of the differences between us and some of the stricter AI safety people is that we believe that the world should accept some bad things happening for the benefits of this technology and people having the agency [to use it].” He then offers examples of what he describes as those “bad things,” beginning with what he frames as minor failures — such as Google’s “AI Overviews” telling users to put non-toxic glue on pizza so the cheese would be less likely to slide off — before turning to harms he describes as more serious.

The first serious harm Bayoumi details is the suicide of Adam Raine, a 16-year-old student whose family has sued OpenAI. According to the column, Raine initially used ChatGPT for homework help before taking his own life, with the family’s lawsuit alleging the chatbot encouraged the act. OpenAI has claimed that Raine misused the platform, the column notes. Bayoumi writes that OpenAI faces additional lawsuits from families alleging harms related to suicide and self-harm.

Bayoumi then turns to military applications of the technology, citing a February 2026 incident in which two Tomahawk missiles struck the Shajareh Tayyebeh elementary school in the Iranian city of Minab on the first day of the war in Iran, killing at least 120 children. A UN fact-finding mission has concluded the United States was likely responsible for the strike, according to the column. Citing a Bloomberg News investigation, Bayoumi reports that an AI-linked targeting system called Maven, made by Palantir, was used in the strike. Unnamed defense officials told Bloomberg that “inside Centcom, which conducted the US attack, some personnel relied too much on the artificial intelligence embedded in Maven Smart System,” according to the column. Bloomberg also reported that the Defense Department “in the past year … rapidly made Maven a cornerstone of the US military apparatus,” the column adds.

In a second military incident, Bayoumi cites CNN reporting from September that an unidentified AI chatbot used by the US military misidentified a Chinese ship in Middle Eastern waters as carrying components of a nuclear weapons program. US planes took off on an interception mission and military personnel prepared to board the ship before the error was caught, according to CNN’s account. CNN also reported that the analyst who generated the flawed analysis then used AI again to write the report that set the mission in motion, the column states.

Bayoumi links those incidents to the financial stakes of AI development. Citing The New Yorker, the column reports that OpenAI “is reportedly preparing for an initial public offering at a potential valuation of a trillion dollars” and that Altman “may become one of the wealthiest and most powerful people on the planet.” The column lists additional AI-related harms Bayoumi attributes to the technology, including “taking jobs away from people, stealing the intellectual property of others, writing some piss-poor undergraduate essays, spying on you and your license plates and bringing us to the brink of world war three.”

The column’s stated conclusion calls for “real, tangible and enforceable government oversight” of AI and argues that “we must collectively have the right to decide how AI should be used.” Bayoumi writes that, “despite what you may hear, China already regulates its AI companies quite aggressively,” and contrasts that approach with what he describes as a US preference for self-regulation.

Bayoumi highlights the “White House Accord on Super Intelligence,” which President Trump posted on social media on September 29 and which remains visible on Trump’s account and the White House’s official account. The accord is signed by Trump — who is identified in the document as “President of the Unites States” — and the chief executives of Meta, Nvidia, Google, Anthropic, xAI and OpenAI. Trump has directed that the term “artificial intelligence” be discarded in favor of “super intelligence,” according to the column.

The accord asks signatories to “implement robust internal controls and monitor the capabilities and alignment of its models” and adds that “over time, it may make sense to codify these steps into laws or regulations,” according to text quoted in the column. Bayoumi writes that Trump, Meta chief executive Mark Zuckerberg and Nvidia chief executive Jensen Huang believe that AI companies should simply regulate themselves.