While Gemini remains the primary foundational model—the core AI used for building software—for Google’s internal work, this shift marks a significant departure from previous restrictions that largely barred staff from using external coding tools.
This move highlights the intense pressure within the AI coding race as technology giants compete to increase developer productivity.
Despite investing billions of dollars into its own Gemini infrastructure, Google is addressing internal feedback that its models have trailed behind competitors like Anthropic and OpenAI in specialized programming tasks.
By providing access to outside tools, Google aims to help its staff work faster, mirroring a recent decision by Amazon to allow its own employees to use rival coding agents after facing similar internal pushback.
The rollout is managed through a quota system, where engineers receive a limited allocation of Claude access to support specific projects alongside Gemini.
This technical arrangement ensures that while Google continues to prioritize its own products, employees can leverage Anthropic’s high-performance models for complex engineering challenges.
The expanded availability follows Google’s massive financial commitment to Anthropic, which includes plans to invest up to $40 billion in the AI startup.