Sam Altman, the CEO of OpenAI, has been taking public swipes at Anthropic after the Claude maker embarked on a nationwide tour of the United States, engaging ordinary Americans on what it describes as the hard questions surrounding artificial intelligence. Altman's commentary, delivered with visible amusement, has added fresh friction to an already competitive relationship between the two companies.
What Anthropic's Tour Actually Involves
Anthropic's initiative involves traveling across different parts of the country to hold conversations with people outside the usual tech and policy circles. The goal, according to the company, is to understand how everyday Americans think about AI development, its risks, and its potential benefits. It is an unusual move for an AI lab, most of which tend to focus their public engagement on Washington briefings and industry conferences rather than town-hall-style outreach. The effort reflects Anthropic's longstanding position that AI safety cannot be decided solely by researchers and engineers, and that broader public input matters. Given the company's soaring valuation, the stakes of getting public trust right have never been higher.
Key Facts
- Anthropic launched a cross-country public engagement tour across the United States.
- The initiative aims to gather perspectives on AI from people outside traditional tech and policy spheres.
- Sam Altman publicly mocked the effort on social media, drawing attention to the rivalry.
- Anthropic has framed public engagement as a core part of its safety-first mission.
- The exchange highlights growing tensions between the two leading AI developers.
Altman's reaction was characteristically wry. He questioned the premise of asking the general public about highly technical AI decisions, suggesting the exercise was more performative than substantive. His comments landed in a context where both OpenAI and Anthropic are navigating intense scrutiny over how much influence the public should have in shaping AI development. Altman and Anthropic's leadership have clashed in tone before, most notably in debates over AI's impact on employment, where both CEOs have shifted their public positions as commercial pressures mount.
Anthropic's approach reflects a belief that legitimacy in AI development requires more than technical expertise. It requires public accountability.Anthropic spokesperson, paraphrased from company materials
A Rivalry That Keeps Getting Louder
The back-and-forth between Altman and Anthropic is not new, but it has grown more visible as both companies expand their public profiles ahead of what many expect will be significant commercial and regulatory milestones. Anthropic's founders, including Dario and Daniela Amodei, left OpenAI to start the company on safety principles they felt were being underweighted. That origin story has always made the relationship between the two firms complicated. The Amodei siblings were recently named to TIME's 100 Most Influential People, a recognition that underscores how seriously the broader world is taking Anthropic's mission, even if Altman finds parts of it amusing.
Anthropic has also been pushing into policy conversations at the highest levels. Earlier this year, Dario Amodei joined Altman and other AI executives in urging Congress to take biosecurity seriously, specifically around synthetic DNA screening as AI models approach PhD-level biology capabilities. The two companies can cooperate on existential risks while competing fiercely on everything else.
Whether Anthropic's public tour produces anything concrete remains to be seen. Critics might agree with Altman that the most consequential decisions about AI will be made in boardrooms and regulatory agencies, not in community halls. Supporters would counter that building public trust early is exactly the kind of long-game thinking a safety-focused lab should be doing. Either way, the exchange has given both companies another round of earned media, and it has reminded the industry that the contest between OpenAI and Anthropic is as much about narrative and values as it is about model benchmarks.
