After an amusing exchange of words with SpaceXAI’s Elon Musk, Sam Altman has now taken a fresh jab at another competitor – Anthropic. In his latest post on X, Altman dismissed a new promotional video from Anthropic, touting it as what he initially assumed to be satire, but it wasn’t.
The video, which asks Anthropic’s users to pose the toughest of questions to AI, caught Altman’s attention, leading him to call out Anthropic for silently downgrading its service and withholding access.
In the post, Altman went on to say, “i thought this was satire, kept looking for the handle to be spelled c1audeai or something.”
In another follow-up post, he added, “hard questions are great but only if we deem you worthy enough to not silently downgrade you, or even get access at all”.
Altman’s comment comes a day after his heated exchange of words with Elon Musk, who accused the OpenAI CEO of scamming people, with Altman calling out Musk for selling public market investors on “short-term data centres”.
What Anthropic’s video says about AI
The original Anthropic post promotes a cinematic advertising-style video and initiative inviting the public to submit their toughest, most uncomfortable questions about AI. Titled “There’s hope in hard questions,” the 90-second film features real voices from surveys and user conversations raising concerns such as: “Can AI be trusted?”, “Will it benefit the majority of people?”, impacts on jobs and society, misuse risks, and broader philosophical issues about humanity and technology.
Anthropic positioned the campaign as a commitment to transparent AI. The video is based on extensive research, including a 52,000-person US survey and feedback from over 81,000 ClaudeClaude users, promising to publicly track progress on addressing these questions while admitting shortcomings.
The company directs people to claude.com/hard-questions to participate.
Altman accuses Anthropic of hypocrisy, gatekeeping
Altman’s response to Anthropic highlights the ongoing rivalry between the two AI safety-focused labs. Previously, Altman has criticised Anthropic for what he sees as an elitist approach, i.e., serving premium products to wealthier users while allegedly restricting broader access and controlling AI usage (such as blocking certain companies from their coding tools).
This comes at a time when Anthropic is witnessing runaway success with its Claude Fable – an AI model that’s considered among the most powerful ones available to humans today. OpenAI followed suit with its GPT-5.6 models, promising similar levels of AI processing performance.
The timing of Altman’s comments adds extra spice to Altman’s slew of controversies, which reemerged after a while over a renewed public feud with Elon Musk. Over the weekend, Musk accused Altman of scamming, especially in light of an Apple lawsuit against OpenAI alleging trade secret theft. Musk posted jabs like “Scam Altman strikes again” and referenced alleged theft of “open source AI charity” and Apple’s technology.
Altman responded by highlighting benchmarks suggesting that OpenAI’s latest model (5.6 Sol) as potentially the world’s best, adding that “the most reliable way to tell is that elon is obsessed with me again.” Musk retorted about upcoming orbital data centers and parole officer quips.
