OpenAI co-founder Andrej Karpathy joins Anthropic to advance Claude AI research

Andrej Karpathy is heading back to the center of the AI race.

The OpenAI co-founder and former Tesla AI director announced Tuesday that he is joining Anthropic, adding another high-profile name to the growing battle for top AI talent between Anthropic, OpenAI, and xAI.

“I think the next few years at the frontier of LLMs will be especially formative,” Karpathy wrote in a post on X, referring to large language models. “I am very excited to join the team here and get back to R&D.”

Anthropic said Karpathy will begin this week and focus on building a team around Claude pretraining research. That work sits at the core of large language model development, shaping how AI systems absorb knowledge, reason through tasks, and develop capabilities before reaching consumers and enterprises.

The hire lands at a tense moment in the AI industry, where the biggest companies are competing for the same small pool of elite researchers. Anthropic has been on an aggressive run in recent months, fueled by soaring demand for Claude and investor expectations that the company could eventually surpass OpenAI’s private market valuation.

Karpathy’s arrival follows another notable move earlier this month when Ross Nordeen, a founding member of xAI and former Tesla engineer, announced he was joining Anthropic. That same day, Anthropic struck a deal with SpaceX to rent compute capacity at xAI’s Colossus 1 data center in Memphis, Tennessee, CNBC reported.

For longtime followers of Silicon Valley’s AI wars, Karpathy’s move carries extra weight.

He helped launch OpenAI in its early days before Elon Musk recruited him to Tesla in 2017 to lead the company’s computer vision and Autopilot efforts. During his time at Tesla, Karpathy became one of the public faces of the company’s self-driving ambitions and frequently presented Tesla’s progress in AI and neural networks.

His role at both companies resurfaced repeatedly during the Musk v. Altman trial, which concluded Monday with the jury and judge ruling in favor of OpenAI CEO Sam Altman.

One email shown during the proceedings revealed how highly Musk viewed Karpathy at the time.

“The OpenAI guys are gonna want to kill me, but it had to be done…,” Musk wrote after hiring him away from OpenAI.

In another exchange presented as evidence, Musk described Karpathy as “arguably the #2 guy in the world in computer vision,” behind fellow OpenAI co-founder Ilya Sutskever.

The testimony drew renewed attention to the early days of OpenAI, when Musk reportedly borrowed several OpenAI researchers to work temporarily at Tesla as the automaker struggled to deliver fully autonomous driving technology. Years later, Tesla still does not sell a vehicle that can safely operate without a human driver ready to intervene.

From OpenAI to Tesla to Anthropic: Andrej Karpathy Returns to Frontier AI Research

A veteran in the AI field, Andrej Karpathy was among the founding members of OpenAI in 2015. Two years later, Tesla recruited him to lead its Autopilot division, where he helped build the company’s driver assistance software and AI systems.

Karpathy studied under renowned Stanford AI scientist Fei-Fei Li and became one of the most recognizable voices in modern AI research. Over the years, he has spoken extensively about large language models, describing them as a new kind of computer operating system.

During his time at Tesla, Karpathy worked closely with several leading AI researchers, including former OpenAI research chief Bob McGrew. He left Tesla in 2022, briefly returned to OpenAI, and later launched Eureka Labs, an AI education startup focused on teaching through AI-native software.

Now, he is stepping back into frontier AI research at a time when the rivalry among major AI labs is intensifying by the month.

Anthropic is betting that researchers like Karpathy can help push Claude further ahead in a market where talent has become just as valuable as compute, chips, and data centers.