A change to Andrej Karpathy’s personal bio has triggered online speculation that he may have left Anthropic, even though no reliable source or official statement has confirmed that.
Users recently noticed that Karpathy removed company information from his profile. Before the edit, the bio reportedly included the word “Anthropic.” Some people said the change happened yesterday, and that detail quickly fed a new round of exit rumors.
One blogger also cited an unverified claim saying Karpathy had already submitted his resignation and politely thanked Anthropic for the opportunity. At this stage, that remains unconfirmed. The source material explicitly said there is no reliable source or official information proving that Karpathy has resigned.
Questions over the timeline
Not everyone accepted the rumor at face value. Some users argued that Karpathy may have changed his bio as early as May this year. They also noted that two screenshots circulating online show a difference of roughly 900,000 followers, which raises the possibility that the screenshots were taken at different times.
That means the currently shared images are not enough to establish when the profile edit happened, and they do not show that Karpathy has already left Anthropic.
Tibo, who was identified in the source as the head of OpenAI Codex, also appeared to respond and clarify the rumor about Karpathy’s departure. The source did not include further detail on what that clarification said.
Anthropic was already in another public debate
The speculation arrived while Anthropic was already being discussed for a separate reason. The night before yesterday, Jensen Huang posted on X for the first time from his personal account and shared an open letter that NVIDIA had co-signed. The letter said the future of AI needs top-tier closed-source models as well as top-tier open-source models.
The letter quickly drew broad attention. Elon Musk later reposted it in support, adding to the intensity of the debate around open-source AI in Silicon Valley. More companies kept joining the signatory list, and OpenAI was among them, but Anthropic did not appear on that list.
Some people linked that development to the new speculation around Karpathy. Still, the source provided no evidence that the two were directly connected.
Karpathy has remained a frequent topic since joining Anthropic
Discussion around Karpathy has continued almost nonstop since he announced that he was joining Anthropic.
On May 19 this year, Karpathy said on X that he had joined Anthropic. He said he would work on the company’s core pre-training team under Nicholas Joseph. As an OpenAI co-founder and former head of AI at Tesla, his move was widely seen at the time as a notable win for Anthropic in the competition for talent with OpenAI.
June restrictions added another layer of attention
By mid-June, the U.S. Department of Commerce had imposed export restrictions on Anthropic’s latest models, Claude Fable 5 and Mythos 5, barring foreign nationals from accessing them. That quickly led to another claim spreading online: that Karpathy, who was described in the source as not being born in the United States, might not be able to access his own employer’s most advanced models.
That claim also became a subject of heavy discussion online.
Claude Tag repost drew separate reactions
Later, Karpathy reposted Anthropic’s introduction to Claude Tag. According to the source, the feature lets Claude join Slack channels in a way similar to a team member, receive @ mentions, and handle tasks.
That repost triggered another round of commentary. Some people said that Karpathy, once seen as a neutral educator through his YouTube tutorials and open-source work, with little obvious commercial framing, had now started publicly backing his employer’s products.
Based on the information currently available, neither the removal of company information from Karpathy’s bio nor the resignation claims circulating online is enough to show that he has left Anthropic. For now, the matter remains unresolved pending any response from Karpathy himself.