Google DeepMind engineers are allowed to use Claude for coding while the rest of Google is restricted to Gemini, causing internal tensions

Some Google DeepMind employees have been granted access to Anthropic's Claude for coding purposes while the rest of Google's engineering staff is restricted to the company's internal Gemini models, a divide that has caused friction as Google simultaneously pressures more employees to use AI and ties AI usage to performance reviews. When the question of equalizing access came up internally, one proposed solution was to remove Claude access from everyone, but DeepMind objected so strongly that several engineers reportedly threatened to leave over it. The tension spilled into public view last week when programmer Steve Yegge posted on X claiming Google's internal AI adoption lagged badly, prompting a sharp public rebuke from Google DeepMind CEO Demis Hassabis, who called the post “completely false and just pure clickbait.”