Two prominent artificial intelligence researchers from Google are planning to join Anthropic PBC, a rival AI firm, underscoring ongoing talent shifts that could challenge Alphabet Inc.’s leadership in the AI sector. Jonas Adler and Alexander Pritzel, key contributors to Google’s Gemini AI model, are set to move to the Claude maker, according to sources familiar with the matter who requested anonymity due to the sensitive nature of the information.

Adler contributed to Google’s AI coding projects, while Pritzel was involved in training AI systems. Their departures follow recent exits of other high-profile AI figures from Google, including Nobel laureate John Jumper, known for his work on protein folding using AI, who is also slated to join Anthropic, and Noam Shazeer, a leading researcher who is heading to OpenAI.

These moves come amid intense competition in the AI industry, as startups like Anthropic and OpenAI edge closer to public offerings, presenting lucrative opportunities for talent to join prior to initial public offerings. Google’s struggles to maintain dominance have been further complicated by internal shifts in resource allocation, with some employees reportedly frustrated over computing power being reassigned to other teams, such as DeepMind’s London-based group focused on pre-training AI models.

Before Shazeer’s departure, one of his projects lost dedicated computing resources to a DeepMind team in an effort to improve cross-team collaboration and streamline AI development. Shazeer had been co-leading Google’s Gemini AI model and working on a novel AI architecture rooted in the transformer technique he helped develop in 2017, which continues to be foundational in AI advancements. Nevertheless, his tenure at Google was marked by both significant acclaim and internal controversy, including employee divisions over his comments on transgender identity and the Gaza conflict.

Jumper’s planned move to Anthropic is expected to be delayed by enforceable non-compete clauses under British law, given DeepMind’s U.K. headquarters. Industry analysts note that engineers from DeepMind are much more likely to move to Anthropic than vice versa, illustrating Anthropic’s success in attracting talent from its larger competitor.

Anthropic itself is expanding beyond core AI research into applications in life sciences and healthcare. The company recently secured a funding round valuing it at $965 billion, surpassing OpenAI and positioning it for a potential public debut as early as this fall.

Google, through a spokesperson and DeepMind CEO Demis Hassabis, maintains confidence in its ability to attract and retain top AI talent despite the competitive environment. Hassabis described the current market as the most fiercely competitive in the tech industry’s history, but emphasized Google’s large and diverse team of researchers as a competitive advantage. Both Google and Anthropic declined to comment directly on the individual departures.