Andrej Karpathy says he remains at Anthropic, denying departure rumors

Why it matters

Karpathy’s visibility makes any change to his affiliation a proxy for leadership shifts at Anthropic, a lab that influences the competitive dynamics of the large‑language‑model market.

Andrej Karpathy said on X that he remains at Anthropic and that rumors of his departure are false, rebutting speculation that began circulating on July 25, 2026.

The speculation centered on a 34-tweet thread by am.will (@LLMJunky), which opened by asking whether Karpathy had removed "Anthropic" from his bio or whether the affiliation had never appeared there. A screenshot showed his profile without the affiliation, but it lacked a timestamp establishing that a change had occurred.

Replies offered conflicting accounts. AnkOnCD (@SVofMike) asserted that the affiliation had previously appeared, while other participants cautioned against treating the screenshot as evidence that Karpathy had departed.

The thread then amplified a more consequential claim: an explicitly labeled "UNVERIFIED RUMOR" that Karpathy had tendered his resignation roughly five hours earlier. The post said he had resigned "calmly and politely," but provided no supporting evidence. Other users repeated the claim without adding independent confirmation.

Karpathy's subsequent comment directly rejected that account. He said he is still at Anthropic and characterized the departure rumors as false. The profile screenshot did not establish either that his bio had changed or that his employment status had changed.

Karpathy's standing in AI research gave the rumor unusual visibility. He previously led Tesla's Autopilot AI team and worked at OpenAI before joining Anthropic in 2024 as a senior research scientist. His response closes off, absent contrary evidence, speculation that the AI lab had lost a prominent member of its technical staff.