The company, which owns Facebook, Instagram and WhatsApp, is pouring unprecedented resources into its AI division in a bid to regain its edge in the fast-evolving race to dominate next-generation technology.
According to industry insiders and recruitment data, Meta is offering total compensation packages worth over £750,000, and in some cases exceeding £1 million, to lure elite AI talent, particularly those with expertise in large language models (LLMs), infrastructure, and generative AI systems.
These offers typically include generous base salaries, stock awards and performance bonuses, making them among the most lucrative in the tech industry.
The move comes as Meta CEO Mark Zuckerberg declares AI the company’s single biggest investment area, overtaking the metaverse.
“Our biggest investment is in advancing AI and building it into every one of our products,” Zuckerberg said earlier this year. The company is developing its own open-weight LLMs under the LLaMA (Large Language Model Meta AI) series, with LLaMA 3 released in April and LLaMA 4 already in progress.
Analysts say Meta’s talent strategy is designed to both accelerate product innovation and avoid falling irreversibly behind key competitors.
In addition to its aggressive hiring tactics, Meta is investing heavily in compute infrastructure. The company is reportedly on track to deploy 600,000 GPUs by the end of 2025, giving it one of the largest AI training capabilities in the world.
This is seen as essential to both training cutting-edge models and attracting engineers who want access to the best tools and resources.
The fierce competition has sparked concern about rising salaries and a brain drain from both academia and smaller startups.
Meta has been actively recruiting from top universities and rival labs, sometimes offering compensation that far exceeds what researchers could earn in the public sector. For academic institutions, keeping top AI researchers is becoming increasingly difficult.
Meanwhile, Meta’s open-source strategy is gaining attention. Unlike OpenAI’s closed model development, Meta has released its LLaMA models to the research and developer community, positioning itself as a champion of transparency and open innovation.
This approach has resonated with some in the AI field, but critics question whether open-sourcing highly capable models could pose security and misuse risks. And while Meta has made notable progress, regaining a leadership position in AI will be a long and uncertain journey, especially as its rivals continue to advance at pace.
RELATED STORIES
