Meta's AI Strategy: Costly Open-Source Push Amid Valuation Concerns

What happened

Meta Platforms has intensified its AI investments in 2025, explicitly sacrificing near-term margins to secure long-term control over AI technologies through open-source initiatives like Llama models. This strategic shift aims to build a dominant AI ecosystem, but it comes with severe financial strain, including a projected $70-72 billion capex surge in 2025 and ongoing Reality Labs losses exceeding $17 billion annually. While the Family of Apps segment shows robust ad revenue growth, the stock's elevated valuation—trading around 28x earnings—already prices in successful AI monetization, leaving minimal margin of safety. Execution risks, regulatory overhangs, and competitive pressures from rivals like TikTok further cloud the path to profitability. Investors must scrutinize whether these heavy bets can yield sufficient returns to justify the current premium.

Implication

Meta's open-source AI push could strengthen its ecosystem moat over time, but the immediate financial burden from soaring capex and Reality Labs losses dampens free cash flow and earnings visibility. The stock's high multiples reflect optimistic assumptions about AI-driven growth, yet regulatory threats and execution missteps could derail monetization efforts. Monitoring ad revenue resilience, capex efficiency, and RL loss trends is essential to gauge whether AI investments are translating into tangible value. Given the limited margin of safety and persistent overvaluation, trimming or underweighting positions may be prudent until clearer evidence of sustainable returns emerges. Overall, the risk-reward skew remains unfavorable, reinforcing a critical stance despite Meta's strategic ambitions.

Thesis delta

The news reinforces Meta's strategic focus on AI and open-source ecosystems, but it does not alter the core thesis of overvaluation and high investment risk. The 'POTENTIAL SELL' stance remains unchanged, as the financial burdens and regulatory uncertainties highlighted in the DeepValue report continue to outweigh potential long-term benefits. No material shift in investment outlook is warranted until evidence of improved capital allocation or reduced downside risks emerges.