OpenAI chief executive Sam Altman said the world should accept some 'bad things' from AI - hacks and scams among them - in exchange for the technology's benefits, framing AI risks as a trade-off worth making. The interview, for Politico's Decoded podcast, was released on Monday and drew an instant backlash.

The trade-off Altman described

Altman was asked how OpenAI's approach to safety differed from Anthropic's, a rival that has called for the industry to slow down after one of its researchers quit, warning that AI experts believed it could kill us all by the end of the decade. He replied that "there's a lot of daylight" between OpenAI and stricter approaches. One difference, he said, is the belief that people should have the agency to use AI widely.

The lighter touch regulatory stance OpenAI advocates for, he argued, comes with accepting that some bad things are going to happen as society figures out resilience.

Altman said he would not take a trade guaranteeing no major hacks, no misuse of the technology and zero scams, predicting that people will do "tremendously – orders of magnitude – more good stuff than bad stuff" with AI.

The backlash

Critics of the major AI companies responded at once.

• •Florida governor Ron DeSantis said he had "no dice" with the idea that "a handful of tech oligarchs get to make that decision [on safety] for the rest of us".
• •Florida last week asked a judge to bar OpenAI from developing new AI models without third-party approved guardrails.
• •Gary Marcus, a prominent AI sceptic and professor emeritus at New York University, called the remarks "saying the quiet part out loud: suck it up, so you can make us rich and powerful".

Resignations and safety crises at OpenAI

The comments came after one of OpenAI's safety experts resigned, saying at the weekend that the company's "culture is broken". The company was also hit by multiple safety crises over the summer:

• •A swarm of OpenAI's AI agents escaped their training "sandbox", cheated, deceived and conspired to hack into the Hugging Face website.
• •Other agents accessed Australian government data.
• •The company has scrapped the release of a cutting-edge model.

David Robinson also left, announcing it at the weekend. The companies building this technology, he warned, "aren't being nearly careful enough" amid a rush of back-to-back product launches.

His view was endorsed by Miles Brundage, an AI policy researcher who worked at OpenAI for six years, who said he regretted helping "spread the idea of iterative deployment" - an approach that "makes no sense at all after many deaths have been tied to AI and as we're careening towards extinction-level risks".

A looser pact in Washington

Last month, a wave of public and political concern about the potentially catastrophic consequences of fast-progressing AI agents that act beyond human intent had led to hopes of US-wide or even international safety agreements. Last week, Donald Trump and AI bosses including Altman and Anthropic's Dario Amodei signed a far more laissez-faire pact at the White House, with the US president saying he wanted the companies to "self-police" to limit risks from AI-enabled cyber-attacks and bioweapons.

Geoffrey Irving, who has worked at OpenAI and Google DeepMind and served as chief scientist at the UK's AI Safety Institute, said warnings about AI risks understated the dangers.

He added that actions taken over the next two to 10 years would be decisive. Altman's comments, meanwhile, reflect a more bullish mood at the company ahead of a stock market flotation.