Joe Maring / Android Authority
TL; DR
- A new report claims that Google considers Anthropic’s latest Claude Code models to be the best of comparable Gemini models.
- Google is forming a new “strike team” of researchers and engineers amid a new push toward using an in-house AI coding tool.
- The company uses AI for about half of its code as of February 2026, while Anthropic uses AI for almost all of its code.
The strike team is said to be led by Sebastian Borgeaud, a Google DeepMind research engineer who previously led preparations for DeepMind. Other notable contributors include Koray Kavukcuoglu, chief technology officer of DeepMind, and Sergey Brin, co-founder of Google. Information says Google DeepMind engineers believe Anthropic’s latest Claude Code models outperform similar Gemini models, leading to the company’s latest push.
Don’t want to miss out on the best Android Authority?
A leaked internal memo passed by Google co-founder Brin reinforced the urgency of the strike team’s mission. Information. “To win the final sprint, we urgently need to close the gap in agent execution and transform our models into core developers,” Brin wrote. To do this, Google plans to create artificial intelligence coding models specifically for internal use and double down on agent functions.
While the Gemini, DeepMind, and Google Cloud teams typically build AI coding models for third-party users and clients, it has a new focus on coding tools for internal deployment. The built-in Google AI coding models are said to be trained on a dedicated Google codebase. Using a private Google code base can improve internal performance, although models trained with Google code cannot be made public.
The move comes as big tech companies, including the aforementioned Spotify, as well as direct AI competitors like Anthropic, rely on coding agents for development. Anthropic’s Boris Cherny, who heads Claude Code, said earlier this year that almost all of the company’s code is generated using artificial intelligence. Meanwhile, as the publication noted, Google’s chief financial officer, Anat Ashkenazi, said during an earnings call in February that Google uses artificial intelligence for only half of its development.
Queue includes internal AI coding tool usage tracking, mandatory training sessions and usage requirements for certain tasks. Google’s in-house AI coding tool is called Jetski, and it wants to understand which employees are more aggressively buying into its use.
This was stated by the Google spokesperson Information said in a statement that it has seen “huge adoption” of these tools. “Using them turbocharges our model and the development of AI tools – that’s where we really focus.”
Thank you for being a part of our community. Read our Comment Policy before deployment.
