Billionaire Bill Ackman Bets $2B on Meta's AI Future: Why He Sees Zuckerberg's 'Deeply Discounted' Play

Pershing Square puts 10% of its capital into Meta at a moment when Wall Street is spooked by the company's $135 billion AI spending plan. Here's Ackman's contrarian thesis.

Bill Ackman, the billionaire activist investor known for concentrated, high-conviction bets, revealed Wednesday that his Pershing Square Capital Management has taken a $2 billion stake in Meta Platforms — allocating a full 10% of the fund's capital to Mark Zuckerberg's AI-driven vision at precisely the moment when Wall Street is most skeptical about the company's massive spending plans.

The disclosure, made in Pershing Square's annual investor presentation, represents one of Ackman's largest new positions in years and signals a stark contrarian view on one of the tech sector's most contentious debates: whether Meta's $115 billion to $135 billion in AI-related capital expenditures for 2026 will generate returns that justify the investment.

"We believe Meta's current share price underappreciates the company's long-term upside potential from AI and represents a deeply discounted valuation for one of the world's greatest businesses," Pershing stated.

Betting Against the Crowd

Ackman initiated the stake in November 2025 at $625 per share, when Meta was trading near recent lows amid growing investor concerns about AI spending. The timing reflects Ackman's trademark contrarian approach: buying when others are fearful, particularly when he believes the fear is misplaced.

Meta shares have fallen 16% over the past 12 months, significantly underperforming the Nasdaq and the broader tech sector. The weakness stems almost entirely from a single concern: is Meta spending too much on AI?

In its fourth-quarter earnings report in January, Meta projected 2026 capital expenditures of $115 billion to $135 billion — nearly double the $72 billion it spent in 2025. That staggering increase raised immediate red flags for many investors. Companies don't typically double their capex in a single year unless they're chasing something with enormous upside or making a massive strategic mistake.

Wall Street is divided on which it is.

The Valuation Argument

Pershing Square's core thesis starts with a simple valuation observation: Meta trades at 22 times projected forward earnings. In Ackman's view, that multiple is too low for a company with Meta's competitive position, cash flow generation, and AI-driven growth potential.

Consider the comparison set: - Alphabet: Trades at roughly 22x forward earnings despite facing intense competitive pressure in search from AI and lower margins than Meta - Apple: Commands a 30x+ multiple despite slower growth and hardware commoditization risks - Nvidia: Trades at 40x+ earnings on AI infrastructure demand - Amazon: Earns a premium multiple despite lower margins and intense e-commerce competition

Pershing's argument: the market is correctly pricing in Meta's elevated AI spending (which depresses near-term earnings) but incorrectly discounting the potential for AI to accelerate revenue and earnings growth in the medium to long term.

If AI meaningfully improves Meta's advertising effectiveness, enables new revenue streams, or creates durable competitive advantages, then 22x forward earnings looks cheap. If the AI spending fails to generate proportional returns, the valuation is about right or even generous.

Ackman is betting on the former.

Why Meta Has Unique AI Advantages

Pershing's presentation highlighted Meta as "one of the clearest beneficiaries of AI integration" — a bold claim given that every major tech company is investing heavily in AI. What makes Meta different?

1. Unmatched user access

Meta owns three of the world's most popular communication platforms: Facebook (3 billion users), Instagram (2 billion users), and WhatsApp (2 billion users). These aren't just large user bases — they represent daily engagement at a scale few companies can match. In an AI-driven world, access to users and their attention becomes even more valuable.

2. The social graph

Meta's most durable competitive advantage is the social graph — the detailed map of relationships, interactions, and preferences among billions of people. This asset is nearly impossible to replicate and provides unique advantages for AI-driven personalization, recommendation, and targeting. While other companies have user data, Meta has relationship data, which is inherently more valuable for many AI applications.

3. Advertising as an AI use case

Meta's core business — digital advertising — is one of the clearest near-term beneficiaries of AI advancement. AI can improve ad targeting (matching ads to users more precisely), creative optimization (generating and testing ad variations), and measurement (understanding what's working). These improvements directly translate to revenue as advertisers pay more for more effective ads.

4. Open-source ecosystem strategy

Unlike OpenAI (proprietary) or Google (mostly proprietary), Meta has bet on open-source AI through its Llama model family. This strategy has risks (giving away valuable technology) but also potential advantages. By making Llama freely available, Meta is building an ecosystem of developers and companies that use its models, potentially creating network effects that feed value back to Meta's platforms.

What Meta Is Actually Building

To evaluate Ackman's bet, it's essential to understand how Meta plans to deploy $115-135 billion in AI capex.

The spending breaks into three categories:

Infrastructure at scale: The bulk of the spending goes toward data centers, custom silicon (Meta's MTIA chips), and massive GPU clusters. Meta is essentially building a private cloud comparable to AWS or Azure, but optimized specifically for AI training and inference. This infrastructure will support both Meta's internal AI development and potentially external customers (though Meta hasn't committed to becoming a cloud provider). Llama model development: Meta recently released Llama 4, including a 288 billion parameter "Behemoth" model that ranks among the world's most capable LLMs. The company is investing heavily in training progressively more powerful models, with internal codenames "Project Avocado" and "Project Mango" for next-generation releases. AI-first product innovation: Meta is rebuilding its entire product suite around AI: - Instagram and Facebook feeds powered by AI recommendations - AI-generated content and filters - AI chatbots integrated into messaging - Smart glasses with Llama integration for real-time AI assistance - VR headsets with AI-driven avatars and environments - Advertising tools that use AI for creative generation and optimization

The question is whether these initiatives will generate returns that justify the investment.

The Bear Case Ackman Is Betting Against

To understand the boldness of Ackman's position, it's worth examining what he's betting against. The bear case on Meta's AI spending is substantial:

1. Overinvestment risk: Meta is essentially doubling its capex in one year. History shows that companies that dramatically increase spending on unproven initiatives often destroy value. The infrastructure Meta is building could end up underutilized or obsolete faster than expected. 2. Open-source value leak: By making Llama freely available, Meta is potentially subsidizing its competitors. Startups and rivals can use Llama to build products that compete with Meta without having to invest in foundational AI research. Critics argue Meta is giving away the store. 3. Competitive threats remain: AI spending doesn't solve Meta's existing challenges. TikTok continues to erode engagement among younger users. Apple's privacy changes have permanently impaired ad targeting effectiveness. Regulatory pressure in the US and EU threatens Meta's business model. 4. Unproven new products: Meta's new product categories (smart glasses, VR headsets, AI wearables) remain niche. There's no clear evidence of mass-market demand, and the company has a track record of spending heavily on hardware products that fail to gain traction (remember the Portal smart displays?). 5. AI commoditization: If multiple companies build comparable AI capabilities by spending similar amounts, the technology could become commoditized. In that scenario, the massive investments generate normal returns at best, destroying shareholder value relative to alternative uses of capital (buybacks, dividends, M&A).

These concerns are why Meta has underperformed. Ackman is betting they're overblown.

What Success Looks Like

Several indicators will reveal whether Ackman's bet is paying off:

Near-term (6-12 months): - Revenue acceleration as AI improves ad effectiveness - User engagement metrics (time spent, content consumption) trending up - Early traction for AI-powered products (smart glasses, chatbots) Medium-term (12-24 months): - Margin recovery as AI investments start generating returns - Growing Llama ecosystem (developer adoption, innovative applications) - Evidence that AI is creating defensible competitive advantages Long-term (24-48 months): - New revenue streams from AI products (wearables, digital assistants, potentially enterprise Llama licensing) - Clear ROI on the $115-135 billion in 2026 spending - Stock re-rating as the market recognizes the value created

If these indicators start appearing, Ackman's $2 billion bet could generate substantial returns. If they don't, the position could weigh on Pershing's performance for years.

The Broader Stakes: Big Tech's AI Arms Race

Ackman's Meta bet is also a bet on the broader AI infrastructure buildout. Across Big Tech, companies are collectively spending over $400 billion annually on AI-related capex:

- Microsoft: ~$145B (Azure + OpenAI partnership) - Meta: $115-135B (Llama + social platform integration) - Amazon: ~$110B (AWS AI infrastructure) - Google: $75-90B (Gemini models + search integration)

This represents the largest coordinated technology investment in history. For perspective, the entire global semiconductor industry's combined annual capex is roughly $200 billion.

Is this productive investment or destructive arms race? Ackman is betting on the former — at least for Meta. His view appears to be that Meta's combination of user access, data advantages, and execution capability will allow it to generate outsized returns from AI investments relative to peers.

The Bottom Line

Bill Ackman has put $2 billion and 10% of his fund's capital behind a simple thesis: Wall Street is underestimating Meta's AI-driven future.

He's betting that: - AI will fundamentally improve Meta's advertising business - Meta's social graph provides durable competitive advantages in an AI world - New AI-powered products will create material new revenue streams - The market will eventually recognize this value and re-rate the stock

The next 12-24 months will reveal whether this contrarian bet was prescient or premature. But one thing is certain: when one of the world's most successful investors puts 10% of his fund into a single company, the market pays attention. For Meta, Ackman's vote of confidence provides valuable validation at a critical moment in its AI-driven transformation.

---