Google adjusts internal policies to allow all engineers to use Anthropic Claude

According to Business Insider, Google has adjusted its internal policies to allow engineers across the company to use the Anthropic Claude model, including Anthropic Opus 5, through the internal development platform Antigravity. Google stated that Gemini remains the primary foundational model developed internally by the company, while third-party models like Claude will be provided according to user quotas, mainly for specific scenarios, as a supplement to Gemini.

Previously, access to Claude was mainly granted to the Google DeepMind team and some high-priority project employees. Reports indicate that despite Google launching the Flash 3.8 model, its coding capabilities are still considered by some employees to lag behind those of Anthropic and OpenAI. Meanwhile, Google has also announced plans to invest up to $40 billion in Anthropic.