How Tech Companies are Snapping Up Global AI Talent
These are the strategies scale-ups and Big Tech companies are using to attract globally mobile AI workers
Zain Ali | 19 February 2025
AI workers are the belles of the ball: in high demand but limited supply.
AI is an arms race, after all. Governments and companies everywhere are rushing to develop the fastest, most sophisticated AI tools known to humanity. But the scarcity of skilled labour is their number one bottleneck. Half of all AI positions are vacant, meaning HR teams have their work cut out for them.
But there’s another twist. AI talent is twice as willing to move to another country for job opportunities. In fact, 42% of top-tier AI researchers are foreign nationals working somewhere else. That’s why hiring functions at tech companies are peculiarly internationally minded.
With such scarcity, these companies resort to unusual methods to secure these workers.
Mark Zuckerberg takes it upon himself to personally email the top AI researchers his company is courting. Chinese recruiters, funded by substantial government subsidies, offer eye-watering wages to German AI workers to lure them to the other side of the world. Not to leave out the big tech companies buying whole AI startup teams with crazy amounts of cash. The list goes on.
But what works? And what doesn’t?
To answer this question, we must first ask:
What do AI workers want?
Do they want money, flexibility, or career advancement? The data shows they want all these things – and more.
Look to the winners. Tesla and Inflection AI cater to tech workers’ preference for flexibility with unlimited time off. Senior researchers get massive compensation packages, and companies like NVIDIA provide generous stock options, meaning that even middle managers make millions a year. Startups like Mistral AI offer equity in their early days.
That’s not to mention paid parental leave, on-site laundry and car maintenance, and even healthcare plans that include egg freezing and IV.
AI workers want something else, though, and this ‘something’ sets them apart from most global workers: they want to play.
They are inventors, after all; intellectuals, and obsessives, with the same pioneering spirit as Newton, Tesla, and Einstein. AI workers want to experiment with the most cutting-edge technology out there. When that opportunity fades, they get bored.
Sometimes, a startup’s commercial pivot can cost it the original cohort of researchers who built it. When DeepMind merged with Google Brain in April 2023, researcher morale plummeted. They were used to working on pioneering projects, not products. Within a year, twelve top researchers left to start new ventures.
But commercial growth doesn’t automatically spell retention problems. Anthropic boosted its headcount by 331.25% between 2023 and 2024 all while launching the highly successful Claude AI LLM, which it leases to enterprises. Its non-hierarchical structure, where all technical employees are given the same title, granted its staff the autonomy to pursue their own versatile research interests. The result is an AI product widely regarded as among the best on the market.
Similarly, over the same period, London-based Augogen AI’s $65m funding rounds didn’t stop it from boosting its headcount by 5x in six months.
DeepMind’s experience might have been a growing pain, one of the unavoidable adjustments of maturity. After all, many researchers would still jump at the chance to get on DeepMind’s commercial arc. But it remains salutary. AI workers are passionate inventors. Product-focused companies will flourish when they balance commercial priorities with an internal culture of innovation and give talented, free-thinking staff the room to be themselves.
The next question:
How Do Companies Attract AI Talent?
Historically, AI recruiting started on campuses. DeepMind, for example, recruited all its researchers from leading UK universities. This still holds to this day. Alibaba works closely with Chinese universities, helping them fill roles with fresh graduates. Japan’s RIKEN Centre for Advanced Intelligence, Germany’s Max Planck Institute, and more; the list of academic recruitment grounds goes on.
However, the number of AI specialist graduates is slim pickings compared to the number of roles waiting to be filled. This is why, according to one tech recruiter, the window for recruiting MIT graduates sits at a 2-week mark, most already holding offers from FAANG-sized companies or disruptive, research-heavy startups.
How do most companies cope with such scarcity? They get talent by fair means or foul. They turn to poaching from their competitors, a task taken so seriously that even Mark Zuckerberg contacts prospects personally.
1. Poaching
A source close to DeepMind said it’s normal for recruiters to approach the company’s researchers with the promise of 800k salaries. Indeed, the CEO of Salesforce capitalised on OpenAI’s disgruntled researchers by promising to match their wages and equity if they resigned and joined his team.
In its early days, OpenAI poached the best from big names like Google, Meta, JP Morgan, Apple and Amazon with the attractive triple whammy offer of hands-on time with specialist chips, an infamously experimental research culture, and massive compensation packages and stock options.
Mistral AI, founded by one of DeepMind’s ex-researchers, poached almost half the Meta team responsible for developing its LLaMA model, not only with the promise of working on truly innovative technology but with equity in limited stock options worth millions. This is a significant draw of fledgling AI startups: the opportunity to get in early, and reap the rewards later on.
2. The draw of startups
Startups led by well-known, pioneering AI researchers can capitalise on their name and network. Take Mistral AI, founded by engineers from DeepMind. Like so many other AI startups founded by former DeepMind researchers — the list includes Inflection AI, Glyphic AI and Orbital Materials — it relied on angel investment from former colleagues. This should come as no surprise. AI workers comprise a small, relatively tight-knit network, many of whom know each other. This is why Meta’s employee referral program is so successful: it rewards employees for capitalising on their networks and bringing in talent.
3. Talent-gutting
“Talent gutting” is a term I see more and more of these days — in short, Big Tech companies mass-hiring fledgling startups through backdoor acquisitions, also known in corporate doublespeak as ‘reverse acqui-hires’. Instead of buying them outright — and having to face trust-busting regulatory scrutiny as a result — they buy the license to their tech and mass-hire the key players with fat wads of cash. Google did it to Character.AI. Microsoft paid Inflection AI $650m to mass-hire its team (including the co-founders) and buy the license to its models. Amazon did the same to Adept AI.
It’s an arms race, and armies win by fair means or foul.
How Globally Mobile AI Workers are Changing the World
Make no mistake: the race to develop AI is an arms race, and AI workers are its frontline soldiers. These troops are globally footloose and, as we’ve seen, sometimes rather mercenary in their approach.
Take the US: only 18% of its top AI researchers received their undergraduate degree there. No wonder 80% of the top H-1B employers are tech companies.
The problem these days is political: Chinese citizens account for the vast majority of AI talent in the US. But a new ‘red scare’ is sweeping through Silicon Valley: a rising, palpable fear that Chinese star talent is sharing US AI secrets with their home authorities.
Whether this is political propaganda or not is a separate question, however one thing that is true is that Chinese companies are trying to lure away European and US AI workers with extraordinary sums of money. Huawei recruiters offered AI employees at Zeiss contracts worth three times their annual salaries. They mean business.
The Chinese government is even willing to fund these sly recruiters. South Korea and Taiwan know all too well the risks of letting their AI employees work in China. After a string of cases where AI researchers brought their secrets with them, both countries introduced stringent counteractive measures, including million-dollar fines and lengthy prison sentences. Silicon Valley companies ask their employees to alert them if this happens.
The repercussions are huge. US authorities scrutinise Chinese H-1B applicants more forensically than ever. The wait times are long — some candidates of extraordinary ability wait up to 5 years just to get their green card. It’s got so bad that plenty of Chinese students (who account for a quarter of all STEM students in the US) are reconsidering their overseas education options.
This is true across the West. In the UK, for instance, the Skilled Worker visa scheme is increasingly expensive for both employer and employee, replete with long wait times, third-party fees, and inordinate red tape. The new government has promised to loosen the rules and make it easier to hire top AI talent, as part of its plan to revitalise the UK economy, but only time will tell.
How are AI companies getting around this rising nationalism and culture of distrust? The answer is simple: they register branches worldwide in flowering tech hubs. DeepMind has branches in Tokyo, Bangalore, and Tel Aviv. Mistral AI, headquartered in Paris, also has branches in Frankfurt. OpenAI? Tokyo and Singapore, Brussels and London.
The future, however, is uncertain. Trump has laid out ambitious AI infrastructure projects and the Starmer government has done the same, but how will they meet these ambitions without a globally mobile workforce of AI employees? Elon Musk seems to know the answer: they won’t. That’s why he’s tweeted recently that he will “go to war over this issue”.
The Bottom Line For Your Strategy
AI workers want the freedom to experiment and work on the world’s most cutting-edge technology. They’d sooner take a pay cut than sacrifice a chance to change the world meaningfully.
To them, it isn’t really about money. It isn’t necessarily about compensation or healthcare plans either. It’s about company culture. Give AI workers the tools of the future, the chance to mould the world, and the freedom and time to do it, and they will thank you.
N.B. But give them some equity too, while you’re at it.
