Google’s AI War Tactic, Pay Them But Don’t Let Them Work!

In the rapidly evolving world of artificial intelligence, top talent is more valuable than ever. Major tech companies including Google, Microsoft, OpenAI, Meta, and others—are in fierce competition to recruit and retain the best minds in the field. As AI capabilities accelerate, driven by advancements in generative models, reinforcement learning, and multimodal systems, companies are desperate to hold on to employees who could potentially develop or refine the next groundbreaking algorithm.

One of the most aggressive tactics reportedly used in this war for talent comes from Google’s AI division, DeepMind. According to Business Insider, DeepMind has been paying some employees not to work—specifically, not to work for anyone else. These individuals are subject to noncompete clauses that can last up to a year and are reportedly still receiving their full salaries during this time.

Paid to Do Nothing; Inside the Unusual Strategy

While it might sound like a dream scenario being paid for a year of extended “paid time off” many employees don’t see it that way. For researchers, particularly in a fast-moving field like AI, 12 months on the sidelines can feel more like exile than vacation. The opportunity to contribute to the latest models, publish in top journals, or participate in high-profile projects is effectively cut off. With each month of inactivity, researchers risk falling behind their peers.

Moreover, the psychological impact shouldn’t be underestimated. Being locked out of your chosen field can lead to feelings of professional stagnation and isolation. According to Business Insider, some DeepMind staff have described the situation as frustrating and demoralizing. Even though they’re being paid, they aren’t allowed to fully use their skills or stay actively engaged in the broader AI research community.

Legal Gray Zones and Loopholes

The use of noncompete agreements is not new, but their use in the tech industry—especially in AI has become increasingly controversial. In the United States, the Federal Trade Commission (FTC) moved in 2023 to ban most non-compete clauses, arguing that they stifle innovation, limit worker mobility, and hurt competition. However, the rule doesn’t apply to all countries, and DeepMind’s headquarters in London falls outside U.S. jurisdiction.

This means DeepMind can continue to enforce non compete restrictions under U.K. labor laws. These clauses often prevent former employees from working for a rival company in a similar capacity for a set period, sometimes up to 12 months. In theory, this gives the original company time to protect its intellectual property and sensitive projects. In practice, it can act as a powerful tool to suppress talent mobility.

Reactions from the Tech Community

The AI community has taken notice and not all the feedback is positive. Last month, Mustafa Suleyman, Microsoft’s VP of AI and a DeepMind co-founder, posted on X (formerly Twitter) that he had been contacted by former DeepMind staff “in despair” over the difficulty of escaping their noncompete agreements. These comments, coming from someone who once helped build DeepMind and now works for one of its main rivals, added fuel to the ongoing debate about ethical employment practices in AI.

While companies are within their legal rights to enforce these contracts (depending on jurisdiction), critics argue that this kind of “golden handcuffs” strategy undermines the spirit of open innovation that AI has historically thrived on. Open-source models, collaborative research, and shared knowledge are hallmarks of the field. Locking researchers into months-long silences stands in sharp contrast to that ethos.

A Wider Trend Among AI Giants?

Google is not alone in finding new ways to retain AI talent. OpenAI, Microsoft, and Amazon have reportedly also offered multi-million-dollar packages, stock grants, and accelerated promotion tracks to keep top researchers from jumping ship. But DeepMind’s approach offering full salaries during long non-work periods may be one of the most extreme examples.

Some analysts speculate that this reflects the growing fear of intellectual property leakage. With AI systems like ChatGPT, Claude, and Gemini competing fiercely in the consumer and enterprise space, even small advantages in model performance or efficiency can translate into major business wins. Companies are therefore investing not only in technology, but in the people behind it—sometimes quite literally paying to keep them off the battlefield.

The Talent Dilemma: Stay, Quit, or Wait It Out?

For AI professionals caught in this bind, the options aren’t easy. Staying at a company where you may feel underutilized or creatively stifled can be just as painful as leaving and facing a year-long professional pause. Some employees have reportedly chosen to ride out the non compete period while quietly planning their next move. Others are exploring academic positions, launching startups in unrelated fields, or consulting in adjacent industries.

In some cases, researchers have taken the time off to upskill, write papers, or learn new programming languages essentially treating the period as a sabbatical. But that luxury isn't available to everyone, and it doesn't negate the sense of lost opportunity.

Ethical Concerns: Innovation Versus Control

The broader concern here is ethical. Is it acceptable for companies to essentially “bench” talent no matter how generously compensated—just to prevent them from helping a competitor? While these agreements are often signed voluntarily, many critics argue that the imbalance of power between corporations and individual researchers makes such contracts inherently coercive.

Furthermore, locking up talent may actually slow down progress in the field. As AI moves from research labs into real-world applications in healthcare, education, finance, and more, there’s growing recognition that no single company should monopolize innovation. Stifling competition or blocking talent flows can lead to slower adoption, fewer breakthroughs, and less societal benefit overall.

Looking Ahead: Regulation, Reform, and Resentment

The FTC’s move in the U.S. to ban most noncompetes could set a precedent, especially if other countries follow suit. There are already calls in the U.K. to review employment laws that allow extended non compete periods. If regulation doesn’t catch up, however, these practices may become more widespread.

At the same time, companies may look for alternative strategies—such as stock vesting cliffs, aggressive counteroffers, or internal transfers to retain valuable staff without resorting to noncompete clauses. Culture, too, plays a role. Firms that foster autonomy, transparency, and meaningful work may have an easier time keeping their people engaged and loyal, even without legal constraints.

Conclusion: Talent at the Crossroads

Google’s DeepMind might be leading the way in an unusual talent-retention strategy, but the implications of its actions are far-reaching. As AI continues to reshape industries and societies, the way we treat the people behind the technology becomes just as important as the tools they build.

In the end, innovation doesn’t happen in isolation it thrives in freedom. If companies want to lead the AI revolution, they may need to find ways to compete that don’t involve silencing their own stars.

Writer: Chrycentia Henryana