OpenAI co-founder and former Tesla AI executive Andrej Karpathy has joined rival AI startup Anthropic, strengthening the company’s push to compete more aggressively in frontier model development.
The move was first reported by TechCrunch and later confirmed by multiple outlets including Reuters. Karpathy’s appointment comes as Anthropic expands its investment in large language model development and enterprise AI capabilities.
TL;DR
- Andrej Karpathy has joined Anthropic’s pre-training team.
- He will work under Anthropic pre-training lead Nick Joseph.
- Karpathy is expected to build a new group using Claude to accelerate AI pre-training research.
- The move intensifies the AI talent war between Anthropic and OpenAI.
- Karpathy previously worked at OpenAI and led Tesla’s AI and Autopilot efforts.
Karpathy announced the move through a post on X, saying he was excited to get back to research and development during what he described as a formative period for large language models.
Reuters reported that he started at Anthropic this week and joined the company’s pre-training division, which handles the large-scale model training runs powering Claude’s capabilities.
According to TechCrunch, Karpathy’s role goes beyond traditional model training. An Anthropic spokesperson said he will create a dedicated team focused on using Claude itself to accelerate pre-training research.
Pre-training remains one of the most expensive and compute-intensive stages in AI development, requiring massive infrastructure investments and specialized research talent. Anthropic appears to be betting that AI-assisted research workflows could become a strategic advantage in the race to build more capable frontier models.
Karpathy remains one of the most influential researchers in the AI ecosystem. He was among OpenAI’s earliest members before joining Tesla in 2017, where he led AI and computer vision initiatives powering Autopilot and self-driving systems. He later returned briefly to OpenAI before focusing on AI education through his startup Eureka Labs.
The hire also highlights Anthropic’s growing momentum in the enterprise AI market. Earlier this month, reports suggested Anthropic had surpassed OpenAI in verified business customers based on enterprise spending data.
Topics For More Insights
Anthropic has also been aggressively expanding its ecosystem. The company recently acquired developer tooling startup Stainless, whose software is used by firms including OpenAI and Google.
The broader AI sector continues to witness intense competition for top researchers and engineers. Several high-profile OpenAI executives, including John Schulman, Ilya Sutskever, and Mira Murati, have departed the company over the past two years to launch new ventures or join competitors.
Karpathy’s move is especially notable because of his reputation for combining deep technical expertise with accessible AI education. His lectures, tutorials, and coding explainers have helped millions of developers better understand neural networks and transformer models.
For Anthropic, bringing in a researcher of Karpathy’s stature signals a deeper investment in foundational model research and potentially AI systems that can increasingly assist in their own development.
Join The Discussion