A new schism has emerged within Google. These are the groups that are allowed to use Anthropic’s Claude and those that are not.
In recent months, some Google DeepMind employees have been given access to Claude AI tools for coding purposes, according to three people familiar with the matter.
Google employees are generally prohibited from using tools that are not manufactured by Google or that are tailored specifically for internal use. Meanwhile, Claude has become one of the most popular AI coding tools across the tech industry.
The decision to allow Google DeepMind employees to use Claude has frustrated engineers in other parts of the company, who are only allowed to use Google’s internal Gemini AI models for coding.
Some employees feel Google’s internal model isn’t as good at coding as Claude’s, two of the people said. Giving some employees access to Claude has caused tensions as Google tells more employees it expects to use AI.
Some engineers have been given specific AI goals that will be factored into their performance reviews this year. In some cases, these employees are said to be expected to use AI to not only generate code but also build tools that make processes more efficient.
A Google spokesperson did not respond to a request for comment from Business Insider.
Google keeps engineers on its internal tools for a variety of reasons. Much of our internal infrastructure is custom built. The company also believes that dogfooding, where employees test and use the products it launches for customers, will help it improve faster.
Other technology companies allow their employees to use external AI models. For example, Business Insider previously reported that Meta employees can use Claude internally.
Google DeepMind CEO looks back
Last week, computer programmer and blogger Steve Jaegge posted on X about Google’s adoption of AI. He said he spoke with Google’s directors and was told that the company was slow to adopt AI.
“TL;DR, Google’s engineering appears to have the same AI implementation footprint as tractor company John Deere,” he wrote to X.
The post sparked a lot of discussion and prompted a candid response from Google DeepMind CEO Demis Hassabis. “Tell your friends to do some real work and stop spreading this absolute nonsense. This post is completely false and pure clickbait.”
Yegge posted again on Monday, saying he had heard from several Google employees that corroborated his claims. Yegge also said that while some DeepMind engineers use Claude, other engineers within the company are not allowed to use it.
“When the issue of equal access came up internally, the proposed response was to remove everyone’s Claude. DeepMind opposed this so strongly that several engineers reportedly threatened to leave,” he wrote.
Have something to share? To contact this reporter via email, please specify the following address: hlangley@businessinsider.com Or text us at 628-228-1836. Use a personal email address and non-work device. Here’s a guide to sharing your information securely.