Billionaire Bill Ackman Loaded Up on These 2 AI Stocks. Should You Buy Them Too?

Bill Ackman increased Pershing Square’s Amazon stake by 65% in Q4 2025 and initiated a new $1.8 billion position in Meta Platforms, giving Meta an 11.4% portfolio weight. The article is bullish on both names because of AI-driven upside in AWS, advertising monetization, and Meta’s valuation, but it is primarily commentary rather than new company-reported financial results. The likely market impact is limited to sentiment and positioning rather than a broad price-moving catalyst.

Analysis

The real signal here is not "Ackman likes AI," but that two of the most levered second-order beneficiaries of the AI capex cycle now sit inside a concentrated, quality-first portfolio. That matters because it validates a regime where the marginal dollar of AI spend is still flowing to platform owners with pricing power, not just to model builders; the market should keep rewarding the owners of scarce compute, distribution, and ad inventory over the next 6-18 months. In that framing, AMZN and META are less a beta trade and more a call option on AI monetization converting from narrative to reported margin expansion.

AMZN’s underappreciated lever is not e-commerce growth, but operating leverage from automation and advertising layered on top of a still-mispriced cloud franchise. The key second-order effect is that AI inference demand can soak up capacity faster than the market expects, which supports both AWS utilization and incremental capex rationality; that should compress any bear case built on "capex is too high". The main risk is timing: if enterprise inference adoption stalls, AWS growth can look capex-heavy for multiple quarters before monetization inflects, creating a window for multiple compression.

META remains the cleaner near-term earnings setup because the core ad engine is already funding the AI spend, so the market is effectively paying little for optionality on product extensions. The consensus is still too focused on Reality Labs drag and too slow to credit that AI improves ad targeting, conversion, and creator monetization in ways that can lift revenue per user without much headline user growth. The contrarian risk is valuation discipline: if ad growth decelerates even modestly or AI capex rises faster than revenue conversion, the stock can de-rate quickly despite strong underlying fundamentals.

The broader winner set extends to suppliers of compute, networking, and data-center power infrastructure, while the losers are lower-quality AI names that need multiple expansion rather than cash flow to justify valuations. GSAT is a more speculative adjacent beneficiary: if Amazon pushes harder into satellite-enabled connectivity, it is an asymmetric but low-probability option on future distribution control rather than a core thesis today. In the next 3-12 months, the market is likely to keep rewarding proof of AI monetization rather than AI ambition, so positioning should favor cash-generative incumbents over pure plays.

Key Decisions for Investors

• Long AMZN vs short a basket of high-multiple AI software names for 6-12 months: own the cash-flowing AI infrastructure beneficiary and fade names that depend on future adoption assumptions; target ~2:1 upside/downside if AWS inference monetization shows up in margins.
• Add to META on 3-6 month pullbacks, funded by trimming weaker ad-tech proxies: the stock still offers the best earnings-to-capex conversion in AI, with downside best contained if core ad growth remains mid-teens.
• Initiate a relative-value pair: long AMZN / short a consumer discretionary e-commerce laggard for 6 months, expressing the view that automation + ads will widen Amazon’s margin gap even if top-line growth moderates.
• Use call spreads rather than outright longs in GSAT only as a speculative satellite internet optionality trade, sized small; treat it as a 12-24 month venture-style position with binary partnership risk.
• Watch for a short-window reversal signal in META if capex guides materially above current consensus without commensurate ad monetization; if that happens, reduce exposure quickly because multiple compression can happen in days, not quarters.