News
Altman: For AI's Benefits, Society Should Accept Some "Bad Things" Happening
OpenAI CEO Altman said in a recent interview that society should accept "negative events" like hacking and fraud brought by AI, because the technology's value will "far outweigh" the harm. The remarks drew strong backlash from political and academic circles.
On October 5, OpenAI CEO Sam Altman dropped a heavy line in a public interview: to get AI’s benefits, society should accept some “negative events” that come with it.
The “negative events” he named include hacking, fraud, and technology misuse. His point, roughly: he would not choose a trade where “we must ensure no major hack happens, no technology abuse occurs,” because the enormous value people create with this technology would far exceed its harms — “the former might be orders of magnitude larger than the latter.”
His Disagreement with Anthropic
Asked how OpenAI and Anthropic differ on AI safety philosophy, Altman replied: “There is a big disagreement between us.”
He put the divide bluntly. Unlike OpenAI, some stricter AI safety advocates think differently: OpenAI believes people should decide for themselves how broadly to use AI. A looser regulatory stance means that as society figures out how to defend against risks, it has to accept that “some bad things will happen.”
The timing was delicate. Days earlier, an OpenAI safety expert had resigned and publicly said over the weekend that the company’s “corporate culture is broken.” This site covered that story (“OpenAI Safety Lead Resigns”).
The Pushback
The backlash was immediate.
Florida Governor Ron DeSantis said he flatly rejects “letting a handful of tech oligarchs decide safety questions for all of us.”
NYU professor emeritus and longtime AI industry critic Gary Marcus was blunter. He said Altman was “saying the quiet part out loud: just put up with it, so we can get richer and more powerful.”
Also Warned on Open Models, and Jabbed at Musk
In the same interview, Altman said two more things.
First, he warned that competitors’ open models could trigger “a coming tidal wave of cybersecurity problems.” He said OpenAI can help customers handle the risk. The subtext: open equals risk, closed equals safety.
Second, his war of words with Elon Musk continued. Altman said: “I don’t like bullies, and I think Elon is one.”
Background: The White House Just Signed a Light-Touch Accord
Just last week, President Trump signed a lighter-touch regulatory accord at the White House with several AI company heads, including Altman and Anthropic CEO Dario Amodei. Trump said he wants AI companies to “self-regulate” to limit AI’s use in cyberattacks and bioweapon development. This site covered that accord.
Altman’s interview pushes the accord’s stance one step further: not just self-regulation, but publicly demanding society accept the cost.
OpenAI has reportedly canceled a frontier model release. This site covered the news that GPT-6.1 Astra was halted over safety concerns. On the safety front, OpenAI clearly hasn’t unified its message internally.
What to watch next: whether “accepting bad things” becomes OpenAI’s formal position in regulatory negotiations — and whether regulators buy it.