Someone threw a Molotov cocktail at Sam Altman's San Francisco home on Thursday night. No one was hurt. A suspect was later arrested at OpenAI's offices, where he had been threatening to burn down the building. San Francisco police confirmed the arrest. Altman, writing on his personal blog the following morning, connected the attack to the publication days earlier of a New Yorker profile he described as "incendiary," noting that someone close to him had warned that the article's release "at a time of great anxiety about AI" could make things more dangerous for him. He wrote that he had "underestimated the power of words and narratives."
The New Yorker profile, reported by Ronan Farrow and Andrew Marantz, was based on interviews with more than a hundred people with knowledge of Altman's business conduct. Its central characterisation was that Altman possesses "a relentless will to power that, even among industrialists who put their names on spaceships, sets him apart." The article focused specifically on the period leading to and following Altman's brief ousting from OpenAI's board in late 2023, examining questions about his trustworthiness as the company moves toward an IPO that could value it above a trillion dollars. Altman disputed the characterisation and what he described as the article's framing and selective use of sources, without addressing specific claims in detail.
The incident is striking in context. OpenAI is navigating a pre-IPO stretch that has been turbulent even by the standards of a company that has rarely had a quiet quarter. This week alone brought a Florida AG investigation into the company over the FSU shooting, a stalking victim's lawsuit alleging ChatGPT reinforced her abuser's delusions while ignoring its own danger flags, a Bloomberg opinion piece arguing the IPO value is threatened by Altman's perceived lack of focus, and now a physical attack on the CEO's home alongside one of the most detailed and critical profiles of Altman yet published by a major outlet. The IPO, reportedly targeting a valuation north of a trillion dollars, is expected within months.
The broader pattern is worth noting carefully, not to diminish the seriousness of a physical attack, but because the pressures converging on OpenAI right now are diverse and structural rather than coincidental. The company is simultaneously defending itself against a wave of harm-related lawsuits, managing political scrutiny from multiple state attorneys general, preparing a complex corporate restructuring to enable the IPO, and operating under intense public scrutiny of its leadership at a moment when AI anxiety is genuinely elevated among large parts of the public. Those pressures do not reduce to the New Yorker's portrayal of its CEO's personality, but they are real, and a company absorbing all of them at once faces a different operating environment than one navigating them separately.
Altman's response to the New Yorker profile was relatively restrained by the standards of how technology executives typically respond to critical press. He did not name the journalists or dispute specific factual claims publicly. His framing was focused on the consequences of published narratives in a high-stakes environment rather than on the accuracy of the reporting. Whether that reflects genuine equanimity, careful legal advice, or a recognition that aggressive rebuttals tend to amplify the stories they contest is not possible to determine from outside. OpenAI did not issue a formal corporate statement on either the attack or the profile as of Friday morning.