The hire gives Anthropic a researcher with experience spanning both frontier AI systems and large-scale production deployment. Karpathy helped launch OpenAI as a founding research scientist before leaving to lead Tesla’s Autopilot and Full Self-Driving AI efforts. He later returned to OpenAI in 2023 before departing again in 2024 to start Eureka Labs, a company focused on AI-powered education tools.
Anthropic said Karpathy will work on pretraining research, the compute-intensive process responsible for building Claude’s foundational capabilities and knowledge base. The company also plans for Karpathy to help build a team focused on using Claude itself to accelerate AI research workflows.
The move arrives during a period of intensifying competition between Anthropic and OpenAI, as both companies push to develop more capable AI systems while expanding commercial products around them. Anthropic has gained momentum recently through growing interest in Claude Code and Cowork, tools that have helped raise the company’s profile among developers and enterprise users.
Karpathy has become one of the most widely followed voices in AI research circles, known not only for his technical work but also for explaining complex machine learning concepts to broader audiences. He popularized the term “vibe coding” and regularly publishes long-form posts about AI development that attract significant attention across the industry. His educational work also includes the online course “Neural Networks: Zero to Hero” and a YouTube channel focused on large language models and neural network training.
The hire reflects how aggressively leading AI firms are competing for experienced researchers capable of bridging theoretical model development with real-world systems engineering. Karpathy’s background at Tesla, where he managed computer vision and autonomous driving programs at scale, gives Anthropic additional expertise in deploying and optimizing large AI systems beyond pure research environments.
Anthropic has also continued expanding its safety and security operations. Separately, the company recently hired cybersecurity veteran Chris Rohlf for its frontier red team, which focuses on stress-testing advanced AI systems against potential threats and misuse scenarios. Rohlf previously worked at Yahoo’s cybersecurity unit known as “The Paranoids” and later spent several years at Meta.
The additions come as Anthropic and OpenAI’s rivalry has become increasingly public. OpenAI CEO Sam Altman recently accused Anthropic of contributing to hostility toward him following an attack on his home, adding another layer of tension to the increasingly competitive AI landscape.
Karpathy’s decision to join Anthropic is likely to draw attention across the industry because of his reputation as one of the few researchers with deep experience in both cutting-edge AI research and large-scale deployment. As AI companies race to improve model performance, reduce training bottlenecks, and build more capable agent systems, talent with that combination of experience has become increasingly valuable.
This analysis is based on reporting from Business Insider.
Image courtesy of karpathy.ai.
This article was generated with AI assistance and reviewed for accuracy and quality.
