OpenAI chief Altman says AI benefits justify risks, warns on overregulation
Version 2: New coverage adds a tech outlet's detailed account of OpenAI's regulatory positioning — its contrast with Anthropic and its backing of California AI safety laws and a US House proposal on external safety evaluators — and an official outlet's second report that Altman warned about the consequences of excessive oversight of AI.
In short
Sam Altman, the head of OpenAI, which develops ChatGPT, said the benefits of artificial intelligence justify the risks and negative events that come with it, according to RBC, TASS and RIA Novosti, which attributed the remarks to an interview with Politico. He said he is not prepared to accept “truly catastrophic risks”, including the possibility of losing control over AI. A fuller tech-outlet account adds that Altman wants AI to stay widely accessible and contrasts OpenAI's stance with Anthropic's stricter approach to regulation.
Read the full story 2 min read
Sam Altman, the head of OpenAI, which develops the ChatGPT chatbot, said the benefits of artificial intelligence justify the risks that come with it and the negative events that occur, according to RBC, TASS and RIA Novosti. RBC and TASS attributed the remarks to an interview with Politico; RBC named Altman in a second report after initially referring to him only as the head of OpenAI, and RIA Novosti's first report carried the headline statement that the benefits of using AI justify the risks. [ 1 , 2 , 3 , 4 ]
Altman said he was not prepared to accept “truly catastrophic risks”, TASS and RBC reported. RBC said he described catastrophic risks, including a serious loss of control over artificial intelligence, as “unacceptable”. [ 1 , 3 ]
The tech outlet 3DNews, which carried a fuller account, reported that Altman said AI should remain widely accessible to the public and that this is the fundamental difference between OpenAI and its competitor Anthropic, which takes a stricter approach to regulating the development of AI systems. Altman told Politico: “We believe the world must come to terms with the fact that some bad things will happen, for the development of this technology and to give people the opportunity to make decisions for themselves,” 3DNews reported. He called it a completely “unacceptable compromise” to agree to a proposal under which a single company would work on AI, control its deployment and decide who benefits from it, the outlet said. [ 6 ]
3DNews also reported that Anthropic had long been regarded as the industry's most determined supporter of AI regulation, but that disagreements between the two leading companies have eased in recent months as OpenAI backed stricter restrictions following high-profile hacking incidents and fraudulent behaviour by its models. It said Altman agreed with a call by Anthropic head Dario Amodei to slow development of the most advanced AI models, that OpenAI followed Anthropic in approving new California laws setting stricter AI safety requirements, and that OpenAI lobbyists recently backed a US House of Representatives proposal requiring leading AI companies to bring in external safety evaluators. [ 6 ]
RIA Novosti later published a second report saying the head of OpenAI warned about the consequences of excessive oversight of AI. The document available to us contained only that headline, with no article text or summary. [ 5 ]
Hacking and fraud featured in the coverage of the risks: RBC said Altman considers the benefits of AI to justify some risks, including hacking and fraud, while 3DNews quoted him saying he would not promise that there would be no serious hacks, no abuse of the technology and no fraud. The documents do not say when or where the remarks were made. [ 1 , 6 ]
Why it matters
The remarks come from the head of one of the most prominent AI developers, whose positions are widely read as a signal of how the industry frames the trade-offs of deploying the technology. The added coverage shows OpenAI positioning itself between calls for tighter regulation and warnings that excessive oversight carries costs. The documents do not indicate any policy decision or regulatory consequence in Russia or elsewhere.
Key facts
- Sam Altman is the head of OpenAI, which develops the ChatGPT chatbot. [ 2 ]
- Altman said the benefits of artificial intelligence justify the risks that have to be taken and the negative events that occur, according to Politico. [ 1 , 2 , 3 , 4 ]
- He said he is not prepared to accept “truly catastrophic risks”, including the possibility of losing control over AI. [ 1 , 3 ]
- RBC reported that he described catastrophic risks such as a serious loss of control over artificial intelligence as “unacceptable”. [ 1 ]
- 3DNews reported that Altman said AI should remain widely accessible and that this is the fundamental difference between OpenAI and competitor Anthropic, which takes a stricter approach to regulating AI development. [ 6 ]
- 3DNews reported that Altman agreed with Anthropic head Dario Amodei's call to slow development of the most advanced AI models, and that OpenAI backed new California laws with stricter AI safety requirements and a US House proposal requiring leading AI companies to bring in external safety evaluators. [ 6 ]
- RIA Novosti published a second report saying the head of OpenAI warned about the consequences of excessive oversight of AI; the document available carried only the headline. [ 5 ]
Confirmed by several sources
- Sam Altman, the head of OpenAI, which develops ChatGPT, said the benefits of artificial intelligence justify the risks and the negative events that come with it. [ 1 , 2 , 3 , 4 ]
- Altman said he is not prepared to accept “truly catastrophic risks”, including the possibility of losing control over artificial intelligence. [ 1 , 3 ]
- Hacking and fraud feature in the coverage of the risks: RBC said Altman considers the benefits of AI to justify some risks including hacking and fraud, and 3DNews quoted him saying he would not promise there would be no serious hacks, no abuse of the technology and no fraud. [ 1 , 6 ]
Still unclear
- When, where and in what format Altman made the remarks. The documents say only that he spoke to Politico, without a date, place or interview format.
- What consequences of excessive oversight of AI RIA Novosti referred to. RIA Novosti's second report on the topic carries only a headline, with no article text or summary in our documents.
- What proposal Altman was rejecting when he called it an “unacceptable compromise” for a single company to work on AI, control its deployment and decide who benefits. Only 3DNews reports this passage, and it does not identify the proposal.
- The rest of Altman's quote about not promising the absence of serious hacks, abuse, fraud and other problems, which is cut off mid-word in the 3DNews text. The document's article text ends mid-sentence, so the full statement is not available.
- OpenAI's reported support for stricter rules in California and in the US House, which sits alongside Altman describing OpenAI's own stance as softer on regulation. This account comes from a single outlet, 3DNews, and the apparent tension within it is not explained in the documents.
What local media are saying
Timeline, local time
- RBC publishes its first report, saying the head of OpenAI urged the world to accept “bad things” for the benefits of AI and called catastrophic risks “unacceptable”. [ 1 ]
- RBC publishes a second report naming Sam Altman and identifying Politico as the outlet he spoke to. [ 2 ]
- TASS reports that Altman said the benefits of using AI justify the risks, and that he is not ready to accept catastrophic ones. [ 3 ]
- RIA Novosti publishes a report headlined that the head of OpenAI considers the benefits of AI to justify the risks. [ 4 ]
- RIA Novosti publishes a second report saying the head of OpenAI warned about the consequences of excessive oversight of AI. [ 5 ]
- 3DNews publishes a fuller account, including Altman's remarks on accessibility, OpenAI's contrast with Anthropic, and the California and US House regulatory items. [ 6 ]