OpenAI CEO Sam Altman has given a wide-ranging two-part video interview to Vanity Fair, published on October 5, 2026, covering President Donald Trump, the midterms, the risk of human extinction from AI, Xi Jinping's state visit and the question of whether the industry should police itself or accept government rules.
The conversation, conducted by Vanity Fair's Mark Guiducci with portraits by Annie Leibovitz, is tied to the magazine's New Establishment list, where Altman takes the top spot. Vanity Fair notes that OpenAI has an agreement with Condé Nast allowing its content to appear in OpenAI search results for a limited term.
Altman on Trump, the midterms and politics
According to Vanity Fair, the second part of the interview puts politics at the centre: Altman discusses Trump, the midterms and the political donations made by OpenAI president Greg Brockman.
Altman is also asked about his meetings in Washington, including a visit to Trump's ballroom, and about his public calls for AI regulation. Vanity Fair reports that even as he asks to be regulated, he accepts it is unlikely to happen, telling the magazine: "We're going to make…I'm going to make many individual decisions that you're clearly not going to like."
AI extinction risk, Astra 6.1 and safety
The interview returns repeatedly to the risk that AI poses to humanity. Altman admits there is a "non-zero" chance this is humanity's last chapter, and the discussion covers alignment, safety breaches and the tension between self-policing and government oversight.
Altman also confirms that OpenAI delayed the release of the latest ChatGPT model, Astra 6.1, after it failed to meet safety standards. He describes current systems as being in a "seriously capable era", where decisions that once looked routine now require far more caution.
The conversation touches on ChatGPT's role in mental health and suicide prevention, a topic covered in Part 1 of the interview alongside Altman's views on Elon Musk, Mark Zuckerberg, pronatalism and preparing for a possible apocalypse.
