The TL;DR
Andrej Karpathy, one of OpenAI’s original 11 founders, joined Anthropic’s pre-training team. He will form a new team using Claude himself to speed up the most expensive phase of the frontier model development.
Andrej Karpathy, one of the original founders of OpenAI and among the world’s most respected AI researchers, announced Monday that he has joined Anthropic. The move is a significant talent coup for Claude’s maker as it scrambles to stay on the cutting edge of language model development.
Review by Andrej Karpathy X, source: X
Karpathy joins Anthropic’s pre-training team, led by Nick Joseph, where he will build a new team focused on an incredibly repetitive goal: using Claude himself to accelerate pre-training research. Pre-training, the massive computational phase that gives the frontier model its basic knowledge and capabilities, is the single most expensive part of building systems like Claude. Finding ways to make that process faster and more efficient could reshape the economics of the entire AI industry.
In an X post that has garnered 13.6 million views, Karpathy wrote that he believes “the next few years on the frontier of LLMs will be especially constructive.” He added that he is “still very passionate about education” and plans to resume that work soon.
TNW City Coworking Space – Where your best work happens
A workplace designed for growth, collaboration, and endless networking opportunities at the heart of technology.
Recruitment encompasses a career arc that has touched almost every major area of modern AI. Karpathy received his PhD at Stanford under Fei-Fei Li, the computer scientist behind ImageNet, specializing in deep learning and computer vision. He was among 11 people who founded OpenAI in 2015, where he worked on deep learning research before leaving in 2017 to join Tesla as its director of AI.
At Tesla, Karpathy led the computer vision teams behind Full Self-Driving and Autopilot, programs that support automakers’ ambitions for autonomous vehicles. He left in July 2022, returned to OpenAI for a year, and then left again in 2024 to found Eureka Labs, a startup that includes AI assistants in education. Eureka Labs’ work is now on hold while Karpathy throws his weight behind Anthropic.
Time is remarkable. Anthropic has emerged as a magnet for high-end tech talent at a time when its biggest competitor, OpenAI, has experienced a series of high-profile departures. In the past two years, OpenAI has lost more than a dozen senior executives and researchers, including CTO Mira Murati, reinforcement learning pioneer John Schulman, and, most recently, three executives who left in one day in April 2026.
For Anthropic, the arrival of Karpathy shows that the company can attract top-level talent as it scales both its research and commercial operations. The company, led by CEO Dario Amodei, has attracted investor interest at a valuation of around $800 billion and is reportedly exploring an IPO that could come as late as 2026.
Karpathy’s new role also emphasizes a broader trend at the frontier of AI: the use of existing models to develop the next generation. If Claude can reasonably speed up his pre-training pipeline, it would be a real sign of iterative self-improvement, one of the skills the AI security community has been eyeing for a long time. Whether that hope pleases or angers observers may depend on how much faith they place in the safety-oriented culture Anthropic has cultivated since its inception.
For now, Karpathy seems to be where he wants to be: back in the lab, building models on the frontier.
