KEY TAKEAWAYS

Reports indicate that Google is compensating some AI researchers for being effectively idle.

Former DeepMind staff find themselves trapped by stringent non-compete clauses.

The primary goal is to guard sensitive projects from being leaked to competitors such as OpenAI or Microsoft.

Recent reports suggest that Google is actually compensating some of its AI researchers to do absolutely nothing. At first glance, this might sound like an enviable position, but there is a significant reason behind this unusual practice—one that goes beyond mere employee perks. This situation has arisen from the enforcement of strict non-compete agreements implemented by Google DeepMind, which is the tech giant’s advanced AI research division.

According to a report from Business Insider, a number of former DeepMind staff members find themselves in a frustrating predicament: they have exited the company but are constrained by legal prohibitions from working with any competing firms for up to a year. During this interim, they continue to receive their full salaries from Google, despite being unable to take on new job opportunities elsewhere.

The rationale behind these non-compete clauses primarily revolves around safeguarding the company’s proprietary interests. The intent is to deter sensitive information about key projects from leaking to competitors like OpenAI or Microsoft. While this approach may serve the company’s interests, it also raises significant concerns regarding fairness and employee autonomy—especially in the rapidly-evolving AI landscape, where innovation demands flexibility and freedom.

Nando de Freitas, a former executive at DeepMind and now holding a VP position at Microsoft AI, recently shared his thoughts on this contentious issue via social media. On X (formerly Twitter), he expressed solidarity with those affected by stating, “Above all, don’t sign these contracts,” labeling them as an “abuse of power.” De Freitas also mentioned that numerous researchers in this situation have contacted him, desperately seeking a way to circumvent the restrictive agreements.

Dear @GoogDeepMinders, first, congrats on the new impressive models. Every week, one of you reaches out to me in despair to ask how to escape your notice periods and non-competes, while also seeking job opportunities because your manager has explained that this is the path to promotion, but…

— Nando de Freitas (@NandoDF) March 26, 2025

DeepMind operates out of the UK, where this practice has sparked significant controversy. Unlike California, where non-compete agreements are prohibited, UK law permits such clauses if deemed “reasonable.” DeepMind claims to enforce these contracts only when absolutely essential to protect critical projects.

Critics of this approach contend that stifling talent—even with financial compensation—ultimately hampers innovation. Keeping highly skilled individuals sidelined not only squanders their potential but also negatively affects their career trajectories and opportunities. It raises a crucial question: Is it ethical for a company to pay employees for doing nothing while preventing them from exploring other paths in a fast-paced field?

This dilemma puts a spotlight on the broader implications of non-compete clauses in tech and AI industries. As these sectors continue to expand and evolve, it becomes increasingly important for companies to balance their proprietary interests with the needs and rights of their talent. If these restrictions lead to employee disillusionment or burnout, the long-term costs could far outweigh any short-term benefits achieved through limiting competition.

In summary, while the concept of being paid to do nothing may initially seem appealing, it belies a troubling reality for many researchers at Google DeepMind. The ongoing situation underscores a critical tension within the tech industry, one that calls for a reevaluation of non-compete agreements, employee rights, and the importance of fostering an innovative environment.