New York – Meta Platforms’ recent strategic moves in artificial intelligence have reignited debates around the company’s future—and whether Mark Zuckerberg’s dual-track growth strategy of “buy and build” can generate lasting returns.
In June, Meta acquired a 49% stake in AI unicorn Scale AI for a reported $14 billion, assigning a valuation of $29 billion to the startup. The deal brought Scale’s founder Alexandr Wang into the fold as head of a newly formed Meta division called Superintelligence Labs. Simultaneously, Meta is pouring record capital into internal development, building AI infrastructure and recruiting top-tier engineers with compensation packages reportedly exceeding $200 million.
On paper, the moves appear bold—aligned with Zuckerberg’s legacy of making transformational bets. Yet in practice, the merger of internal innovation with high-priced external acquisitions is proving difficult to synchronize, raising concerns about strategic clarity, efficiency, and value creation.
Escalating AI Ambitions, Escalating Costs
Meta’s capital expenditures are on track to hit nearly $70 billion in 2025, triple the levels from just four years ago, according to data from Visible Alpha. Despite holding about $40 billion in net cash, Zuckerberg is reportedly exploring up to $30 billion in leveraged financing from private equity giants such as KKR and Apollo Global Management to support Meta’s AI ambitions.
This massive allocation of capital follows several mixed outcomes in past ventures. The company’s pivot to the metaverse in 2021—a move that included rebranding the company from Facebook to Meta—has so far produced meager financial returns. Reality Labs, the division behind the effort, posted a $4.2 billion loss in Q1 2025 alone, adding to over $60 billion in accumulated development costs.
The acquisition of WhatsApp, too, remains an outlier in terms of monetization. Despite 2 billion global users, monetization has lagged due to its end-to-end encryption model and user privacy commitments. Only recently has Meta begun testing ads within the platform—a delicate shift given its origins and philosophical positioning against advertising.
AI Integration: Inspiration or Imitation?
Within the AI domain, Meta is facing stiff competition not only from legacy rivals like Microsoft and Google, but also from rising entities like OpenAI and Anthropic. Despite Meta’s release of LLaMA 4 this past April, internal feedback has reportedly cited a lack of vision and fragmented execution.
Meanwhile, the company has aggressively recruited top AI talent from its competitors, but questions persist about whether leadership can integrate the disparate approaches—external acquisition, internal development, and short-term commercialization—into a unified AI roadmap.
Valuation Reflects Uncertainty
Meta’s market valuation underscores these doubts. The company trades at 16x expected 2025 EBITDA of $109 billion, according to Visible Alpha—below peers such as Pinterest and Reddit. Analysts estimate that Instagram now contributes roughly 36% of Meta’s total ad revenue, worth an implied $65 billion. Facebook, despite steady performance, is growing at half the pace Instagram did between 2018 and 2021.
Collectively, Facebook and Instagram make up over 80% of Meta’s implied enterprise value. That leaves roughly $300 billion in market cap attributed to WhatsApp, Reality Labs, and AI—an underwhelming figure considering the hype and resources poured into these units.
For comparison, Tesla and Nvidia—under Elon Musk and Jensen Huang, respectively—enjoy valuation premiums tied to confidence in visionary leadership and long-term technology plays. Meta, on the other hand, appears to be receiving a more skeptical discount from investors.
A Strategy in Search of Cohesion
Meta’s approach to AI reflects a high-stakes gamble: that aggressive investment and top-tier talent alone will yield breakthrough innovations and future revenue streams. But as with many traditional corporate mergers, combining in-house development with acquired capabilities may not automatically produce synergy. Instead, the effort could end up as a costly mosaic of overlapping efforts lacking coherence.
Until Meta can articulate and execute a clearer AI narrative—one that translates innovation into monetizable products—the market is likely to remain cautious. For Zuckerberg, the challenge isn’t just building AI tools or buying talent. It’s proving that these efforts can be aligned, sustainable, and profitable at scale.
