DeepMind CEO says Meta is behind in AI and scrambling to compete

Meta is not leading the AI race right now

According to Google DeepMind CEO Demis Hassabis, Meta lags in the generative AI race. On the Lex Fridman podcast, he said Meta “is not at the frontier” of AI development, suggesting they are reacting to missed opportunities.

While Meta was early in releasing open models like LLaMA, their reception has been lukewarm. Competitors like OpenAI and DeepMind have captured market leadership and public imagination, pushing Meta into reaction mode.

Meta is offering massive salaries to recruit talent

Meta has launched an all-out hiring blitz, reportedly offering up to $100 million compensation packages to attract AI stars.

These deals include equity grants, bonuses, and long-term incentives aimed at luring talent from labs like OpenAI, Anthropic, and DeepMind.

It’s a high-stakes gamble intended to accelerate Meta’s comeback in AI. The tactic mirrors Meta’s past strategy during mobile’s rise: spend big, hire fast, and outscale rivals. Only this time, it’s with algorithms, not apps.

Meta’s strategy reflects urgency and ambition

Demis Hassabis described Meta’s hiring spree as “rational,” given their current position. Meta appears to be in damage-control mode, launching its Superintelligence Lab and raiding top competitors for talent.

With AI rapidly advancing, Meta’s leadership likely feels the pressure to catch up before falling irreversibly behind.

The company is betting that a sudden influx of expertise will reignite innovation internally and help it deliver transformative AI products capable of challenging ChatGPT or Gemini.

DeepMind’s CEO questions money as the motivator

“There are more important things than just money,” Hassabis said, emphasizing that true innovation in AI often stems from mission-driven researchers.

While competitive pay is necessary, attracting top minds who want to ensure AI is developed responsibly is not always sufficient.

Hassabis implied that the best researchers prioritize ethics, alignment, and long-term impacts over flashy salaries. In high-stakes fields like AI, culture and purpose can outweigh even the most generous offers.

Meta launched a new Superintelligence division

Meta’s new Superintelligence Lab, launched earlier this year, is tasked with developing advanced AI systems akin to AGI. As of July 2025, Meta’s Superintelligence Labs is co‑led by Alexandr Wang (former CEO of Scale AI) and Nat Friedman (former CEO of GitHub).

Underwhelming reactions to earlier LLaMA releases spurred the initiative. With Zuckerberg personally involved in recruitment, the lab symbolizes Meta’s aggressive pivot to remain competitive in a space now dominated by OpenAI and Google.

Former OpenAI staff are jumping ship to Meta

Several OpenAI researchers, including Shengjia Zhao and Shuchao Bi, have accepted Meta’s offers. These hires reflect how intensely competitive the talent market has become.

Meta is cherry-picking individuals who’ve already worked on frontier models, shortcutting years of research by bringing in experts with inside knowledge.

While it’s a strategic coup, it also raises questions about knowledge transfer, IP, and whether culture clashes might hinder the progress they hope to accelerate.

Anthropic says mission beats money for retention

Anthropic cofounder Benjamin Mann shared that many employees reject lucrative offers because they’re deeply aligned with the company’s mission.

Anthropic positions itself as a values-first lab focused on AI safety and alignment. Staffers there reportedly weigh ethical influence more heavily than wealth.

Mann’s comment, “My best case at Meta is we make money, my best case here is we affect the future of humanity,” has become a rallying point for purpose-driven AI professionals.

AI salaries have skyrocketed across the board

From OpenAI to startups, AI compensation has soared. OpenAI technical staff average $292,000 in base pay; Anthropic averages $387,000.

Some new labs, like Thinking Machines, offer salaries above $500,000 for top talent, but these figures exclude equity or bonuses, which can multiply the total comp substantially.

The result? Even early-career engineers can now command seven-figure packages. With talent in short supply and demand skyrocketing, this trend shows no signs of slowing.

AI wasn’t always such a lucrative field

Demis Hassabis reminisced about DeepMind’s early days, when AI research was underfunded and mostly academic. “I didn’t pay myself for a couple of years,” he said, highlighting the contrast with today’s gold rush.

Back then, raising seed capital was a challenge. Today, interns earn what entire labs once raised in funding rounds. The shift underscores AI’s transformation from niche research to one of the most coveted sectors in global technology.

The talent war may inflate AI development costs

As competition intensifies, costs are ballooning. Multimillion-dollar compensation packages, hardware scarcity, and poaching wars stretch even the deepest budgets.

Analysts worry this bubble may drive unsustainable spending and favor only the most prominent players.

If smaller labs can’t keep up financially, innovation could become centralized, slowing the diversity of research and concentrating power among a few mega-corporations like Meta, Google, and Microsoft.

Google claims its AI retention remains strong

Despite Meta’s aggressive hiring, Google CEO Sundar Pichai insists retention is healthy. On a recent earnings call, he said the company closely tracks hiring and attrition trends, which are steady.

Google DeepMind’s prestige and resources may help shield it from the churn. While high-profile exits grab headlines, Pichai suggests that overall talent stability remains intact across Alphabet’s sprawling AI teams.

Meta’s Llama model underwhelmed on launch

Meta’s LLaMA models, though open and accessible, didn’t capture the same excitement as GPT-4 or Gemini. Critics cited lackluster performance benchmarks and a weak application ecosystem.

This underperformance likely spurred Meta’s renewed urgency to hire proven talent. Open-sourcing the model made waves, but failed to translate into meaningful dominance.

With its new lab and mega-hires, Meta is trying to rewrite that narrative and stake a new claim in the AI arms race.

Noncompete clauses are making a comeback

Some companies have reintroduced or strengthened noncompete clauses to guard their intellectual capital. At Google DeepMind UK, employees can be blocked from joining rivals for up to a year.

These contracts aim to prevent knowledge leakage and slow the talent drain. Critics say they stifle mobility and innovation, but for companies racing toward AGI, keeping top minds from joining competitors may be a necessary evil.

Meta’s tactics mirror past Silicon Valley plays

Meta’s current recruitment blitz isn’t new; it’s a playbook we’ve seen before. In mobile and social tech, Meta outspent competitors to acquire talent and speed up development. Back then, it worked.

In AI, however, winning may require more than just engineering talent. Ethical leadership, trust, and innovation credibility could define winners equally. Whether Meta can replicate its past success here remains uncertain.

AI hiring is creating ripple effects in tech

The AI boom is pulling talent from across industries. Fintech, gaming, robotics, and cloud infrastructure engineers are jumping ship for AI roles. This has caused shortages in other fields and may broadly reshape tech companies’ priorities.

Non-AI firms now face retention struggles too, as employees seek higher pay and more futuristic challenges. The talent gravity of AI is warping the entire tech labor market.

Want to see just how fierce the AI talent race is getting? Meta’s latest hires say a lot without saying much.

In AI, being first isn’t always the same as winning

Meta’s race to catch up reminds us that AI isn’t just about who starts first but who adapts fastest and delivers safely. Trust, mission, and long-term strategy matter as much as technical breakthroughs.

DeepMind’s caution highlights that rushing toward AGI without ethical safeguards could backfire. Winning this race will take more than speed or size; it will take vision, responsibility, and sustainable innovation.

Want a glimpse of how high the stakes are? Meta just brought in two more top minds from OpenAI, and the race is only heating up.

What do you think about Meta’s AI race with its rivals being slow? Will Meta catch up and gain more popularity in AI? Please share your thoughts and drop a comment

Read More From This Brand:

• Meta’s AI takes a creative turn with new video editing powers
• Meta teams up with Scale AI in $14.3B deal and a big leadership shakeup
• Meta snatches Apple AI boss in bold move to rival tech giants

Don’t forget to follow us for more exclusive content on MSN.

If you liked this story, you’ll LOVE our FREE emails. Join today and be the first to get stories like this one.

This slideshow was made with AI assistance and human editing.