Meta is escalating its AI talent war with a novel and complex financial maneuver. The company is now offering to buy a minority stake in NFDG, the venture capital firm founded by its latest star hires, Nat Friedman and Daniel Gross.
The deal, reported by The Wall Street Journal, gives the firm’s investors a rare opportunity to cash out at full value. This move comes after Meta failed to acquire several key AI startups, shifting its strategy to poaching top executives instead.
The offer cleverly resolves a major conflict of interest. Friedman and Gross must now cease outside investing to lead Meta’s new superintelligence labs. When asked on X if he would stop investing, Friedman gave a one-word reply: “Yes”.
A Venture Fund Buyout to Seal the Deal
The tender offer proposes to buy out a minority stake from NFDG’s limited partners at the current net asset value, an unusual and attractive offer in a market where secondary transactions often come at a discount. According to people familiar with the situation, Meta could acquire as much as 49% of the firm’s holdings.
To alleviate concerns about information sharing, the deal is structured so Meta will not have information or governance rights in the underlying startup portfolio, which includes companies like AI voice platform ElevenLabs and accounting software Basis.
NFDG will cease making new investments. Its first fund of about $1.1 billion, raised in 2023, has reportedly seen its investments appreciate fourfold. The existing assets will be managed by the remaining team, with guidance from an advisory group including Stripe co-founder John Collison.
A ‘Buy or Poach’ Playbook Forged in Rejection
This intricate financial play is the latest evolution of Meta’s aggressive “buy or poach” playbook. The strategy was forged after repeated rejections on the acquisition trail, including failed attempts to purchase generative video startup Runway and others.
The most high-profile example was its failed bid for the $32 billion startup Safe Superintelligence (SSI). Unable to buy the company, Meta pivoted and hired its then-CEO, Daniel Gross. This pattern underscores a deliberate, multi-billion-dollar campaign to acquire top-tier talent by any means necessary.
The strategy is a direct response to a firestorm of internal challenges. Meta has hemorrhaged talent and faced significant technical setbacks, postponing its ambitious Llama 4 “Behemoth” model. This turmoil fueled what anonymous engineers described as a “panic mode,” with immense pressure on the GenAI organization to justify its costs.
The Talent War’s Ripple Effect: New CEOs and Scrambled Defenses
Meta’s raid on SSI’s leadership triggered immediate consequences. Today, co-founder Ilya Sutskever announced he would take over as CEO, just days after Gross officially departed on June 29. Sutskever emphatically reaffirmed the company’s independent path.
In a post on X, he dismissed acquisition rumors, stating, “You might have heard rumors of companies looking to acquire us. We are flattered by their attention but are focused on seeing our work through.” He projected confidence in his team’s ability to continue its mission, adding, “We have the compute, we have the team, and we know what to do. Together we will keep building safe superintelligence.”
The aggressive poaching also ignited a fierce public and private battle with chief rival OpenAI. In just one week, Meta hired at least eight researchers from the AI leader, sparking a defensive scramble.
A leaked internal memo revealed the depth of the reaction inside OpenAI. Chief Research Officer Mark Chen expressed a raw sense of violation, writing, “I feel a visceral feeling right now, as if someone has broken into our home and stolen something.” The memo detailed plans to “recalibrate comp” to prevent further departures.
The conflict had already spilled into public view. OpenAI CEO Sam Altman claimed Meta was offering massive compensation packages to lure his developers, a charge one new Meta hire, Lucas Beyer, publicly called “no, we did not get 100M sign-on, that’s fake news.” on X.
These frantic moves culminated in Mark Zuckerberg’s July 1 announcement of the new Meta Superintelligence Labs. The formation of MSL formalizes the roles of the poached talent, creating a new, centralized division to accelerate Meta’s AI ambitions.
By acquiring an elite team and now untangling their financial conflicts, Meta has secured a critical short-term victory. However, its chaotic, crisis-driven approach raises serious questions about its ability to build a stable, long-term foundation for AI leadership.
