Sam Altman tells Politico world must accept some AI harms for benefits
OpenAI CEO Sam Altman stated during an interview for the debut edition of POLITICO’s Decoded that the world should accept some bad things happening
to secure the broad benefits of artificial intelligence.
Altman Rejects Frontier Lab Monopoly Claims
Altman addressed criticisms regarding regulatory capture during his interview with POLITICO. He rejected the premise that a single laboratory in San Francisco should control frontier models and dictate the distribution of benefits. I disagree, but I understand the perspective of people who are like, ‘This technology is going to get so powerful, and it’s so dangerous, that a single lab in San Francisco should have it and make sure nothing bad happens,’
Altman told politico.com.
The OpenAI chief called that arrangement an unacceptable trade-off that conflicts with the lighter-touch regulatory approach favored by his company. He noted that he would not accept a guarantee of zero scams, zero hacks, and zero misuse if it meant restricting technological progress, adding that human output with AI will generate orders of magnitude more good than harm. He maintained, however, that his company still rejects catastrophic risks such as a serious loss of control over advanced systems.
These remarks follow Altman’s testimony before the United Nations Security Council, where he warned that humanity could lose control of the future to artificial intelligence. In that UN address, Altman called for urgent government intervention and international coordination. This messaging mirrored his previous appearance before a United States Senate subcommittee, where he warned about dystopian superintelligence.
Anthropic Defends Frontier Model Regulations
Responding to Altman’s remarks, an Anthropic spokesperson told politico.com that its regulatory proposals apply exclusively to frontier models. The spokesperson pointed to a prior social media exchange involving Dario Amodei and an investment manager.
During that August exchange, Amodei defended his company’s policy proposals. Anthropic has always made its policy proposals very carefully,
Amodei stated, according to politico.com. We try very hard to make proposals that disadvantage (slow down) frontier AI companies while advantaging smaller competitors.
DevDay 2026 Developer Announcements and Product Launches
Alongside policy debates, OpenAI launched its annual developer conference at Fort Mason in San Francisco, introducing more than 20 major product announcements across ChatGPT, Codex, and agent platforms, according to openai.com and techiexpert.com. Techiexpert.com reported that opening day admissions cost $650 while the keynote streamed for free.
The conference showcased accelerated capabilities driven by GPT-6 Astra, an internal model that scored 100 percent on the ExploitBench cybersecurity test and 98 percent on FrontierMath Tier 4, per techiexpert.com reporting. Thibault Sottiaux explained that internal engineering gains allowed the company to release a volume of products, including Codex multiplayer programming, live audio generation, and autonomous coding agents.
OpenAI also revealed a ChatGPT Pro Max tier priced at $500 per month. Techiexpert.com noted this plan offers expanded memory, 100 gigabytes of storage, and inference clusters powered by Cerebras hardware that run over 14 times faster than standard tiers. Developer pricing plans were restructured into Free, Prototype, and Accelerate tiers to assist teams testing agent workflows before deployment.
Broader Political and Policy Discussions
The Decoded interview also covered Donald Trump’s new artificial intelligence accord with major technology firms, attempts to rebrand superintelligence, autonomous hacking incidents involving OpenAI models, and industry financial influence in United States politics. POLITICO journalists Brendan, Christine Mui, Calder McHugh, and Sam Sutton produced the launch edition of the newsletter and podcast, drawing on reporting from approximately 50 journalists worldwide.
Altman also discussed the influence of effective altruists within OpenAI and shared perspectives on international political figures, including French President Emmanuel Macron and California Governor Gavin Newsom.