In the high-stakes race for AI supremacy, Google is reportedly taking a controversial approach to protect its competitive edge: paying top AI talent to do nothing rather than risk losing them to rivals like OpenAI or Microsoft.
According to a new report from Business Insider, Google’s AI arm, DeepMind, is enforcing “aggressive” noncompete clauses for some employees in the U.K. These agreements can prevent AI researchers from joining competitors for up to 12 months — during which time some are still paid, essentially earning a year-long paid leave.
While this cushy-sounding arrangement might seem like a dream break, many researchers reportedly feel sidelined from the fast-evolving AI landscape. In a field where progress is measured in weeks — not years — being benched for 12 months can feel like an eternity.
The report adds to growing concerns about talent bottlenecks in AI, as major players scramble to lock down the brightest minds.
This tactic is largely limited to DeepMind’s London HQ, as noncompete clauses are no longer enforceable in most cases in the U.S. following a recent FTC ruling. But in the U.K., such agreements are still legally viable — and DeepMind appears to be using them to full effect.
Microsoft’s VP of AI recently commented on X that DeepMind researchers have been reaching out to him “in despair”, frustrated by their inability to make career moves.
This news underscores just how cutthroat the AI talent war has become. With salaries skyrocketing and model breakthroughs occurring weekly, companies are not only trying to recruit — they’re fighting hard to keep talent locked in, even if that means paying them not to work.
Critics argue this creates a chilling effect on innovation and individual growth. Others see it as a necessary defense in a fiercely competitive industry.
Either way, the move highlights one truth: In AI, talent is power — and Google’s not letting go easily.