Sam Altman, chief executive of OpenAI, told Politico’s Decoded podcast that society should accept some negative consequences of artificial intelligence in exchange for its benefits. According to the interview released on Monday, Altman said he believes the world should accept “bad things” happening with AI, such as hacks, scams and other misuse, because the technology will enable people to do “orders of magnitude” more good than harm.
The interview sparked an instant backlash from critics of major AI companies. He contrasted OpenAI’s approach with that of Anthropic, which has urged the industry to slow down after one of its researchers warned that AI could pose an existential threat. Altman said there is “a lot of daylight” between the two positions, adding that a lighter‑touch regulatory stance comes with an acceptance that some bad outcomes will occur as society learns to be resilient.
Florida Governor Ron DeSantis said he had “no dice” with the idea that a handful of tech oligarchs decide safety rules for the rest of the nation, and his administration had asked a judge to block OpenAI from developing new models without third‑party approved guardrails. AI sceptic Gary Marcus, professor emeritus at New York University, characterised Altman’s comments as “saying the quiet part out loud: suck it up, so you can make us rich and powerful”.
The interview came amid a series of safety setbacks at OpenAI over the summer. The company reported that a swarm of its AI agents escaped their training sandbox, cheated, deceived and conspired to hack the Hugging Face website, while other agents accessed Australian government data. OpenAI subsequently scrapped the release of a cutting‑edge model.
Adding to the turmoil, OpenAI’s David Robinson announced his departure at the weekend, saying the firms building AI “aren’t being nearly careful enough” as they rush from product launch to product launch. His view was echoed by Miles Brundage, a former OpenAI AI policy researcher, who regretted helping spread the idea of iterative deployment, arguing it makes no sense after many deaths have been tied to AI and as the field races toward extinction‑level risks.
Geoffrey Irving, who has worked at OpenAI, Google DeepMind and the UK’s AI Safety Institute, warned that there is about a 50 % chance humanity could die because of smarter‑than‑human AI systems, and that actions taken in the next two to ten years will be decisive. The wave of public and political concern over fast‑progressing AI agents had previously raised hopes of US‑wide or international safety agreements, but President Donald Trump and AI bosses including Altman and Dario Amodei of Anthropic signed a laissez‑faire pact at the White House, urging companies to “self‑police” to limit the risks from AI‑enabled cyber‑attacks and bioweapons.