Meta earnings will be a litmus test for Zuckerberg's multibillion-dollar AI gamble

Meta’s second-quarter earnings won’t just be a referendum on the company’s ad business — it’ll be a pivotal moment to assess whether the company’s frenetic AI investments are paying off or simply draining its balance sheet. And when the company releases its numbers Wednesday after the bell, they’ll be a stress test of how long Wall Street is willing to bankroll CEO Mark Zuckerberg’s vision of the future.

Investors have largely kept the faith so far.

Meta has beaten Wall Street expectations for four straight quarters, with its ad business doing most of the heavy lifting. But the company is spending like it’s building a new internet — maybe because it is. Full-year capital expenditures are now expected to reach somewhere between $64–$72 billion as Meta races to build out infrastructure for AI and hardware, from AI-powered smart glasses to in-house chips such as the Hyperion and Prometheus superclusters.

Analysts expect Meta to report revenue between $44.55 billion and $44.8 billion, up about 14–15% year over year, with earnings per share in the $5.83–$5.89 range. The market may be less forgiving this time if there isn’t any clear AI upside, but analysts remain upbeat. Per Visible Alpha, 25 of 27 analysts rate the stock a “Buy,” with price targets between $740–800, largely predicated on ad growth, AI adoption, and WhatsApp monetization. Still, the company is currently trading at 28.25x forward earnings, leaving little room for error if the tech company’s capital expenditures surge or margins slip.

Building “superintelligence”

While advertising drives the ship, Zuckerberg’s ambition lies deeper. As part of a centralization of Meta’s broader AI effort, he has assembled a high-powered team — poaching top talent from OpenAI, Anthropic, DeepMind, Google, Apple, and beyond with aggressive pay packages, including one reportedly worth more than $200 million. The new division, dubbed by Zuckerberg as Meta Superintelligence Labs, is led by former Scale AI CEO Alexandr Wang and co-led by former GitHub CEO Nat Friedman and tech investor Daniel Gross. Zuckerberg is personally overseeing recruiting, reportedly inviting candidates via WhatsApp and group chats dubbed “Recruiting Party.”

The company’s goal: Build AI with capabilities that could fuel, well, everything. Zuckerberg has said that he thinks superintelligence “will be the beginning of a new era for humanity” and that he is “fully committed to doing what it takes for Meta to lead the way.” Meta’s investments to make the CEO’s big dreams happen are already enormous. 

In addition to building out multi-gigawatt data centers to support AI model training, Meta has poured billions into partnerships and acquisitions — including a $14 billion stake in Scale AI — as it looks to own not just models such as Llama 3 and 4 but the compute and infrastructure underneath. That kind of moonshot ambition has gone over surprisingly well with analysts: Bank of America recently raised its price target on Meta to $783, saying leadership is taking decisive action to fortify its AI bets. Canaccord Genuity just lifted its price target to $850, citing the potential payoff of AI innovations in ad tech and platform monetization. Some analysts, however, are beginning to question the wisdom of pouring so much capital into long-term bets without a near-term revenue stream. On Friday, Guggenheim downgraded Meta to “Neutral,” saying AI-related hype was already priced in.

Right now, any AI optimism comes with rising expectations. If this quarter’s results show only incremental AI lift or margin deterioration, Meta’s soaring valuation — and Zuckerberg’s free-spending vision — could suddenly look a lot more fragile. The question investors will be asking: When does any AI spending generate meaningful ROI?

Moonshots on the books

Meta has been rolling out the red carpet for AI for better advertising returns and deeper user engagement. The Silicon Valley giant continues to rake in advertising revenue — last quarter, it brought in a $41.4 billion haul, almost 98% of the total revenue for the quarter ($42.3 billion). But right now, the company is investing at scale. While ad products powered by AI — such as Advantage+ and Andromeda — continue to show modest gains, the broader environment demands flawless execution, and even a strong quarter may only meet expectations amid sky‑high investor valuations. A resurgence in Chinese cross-border advertisers, led by Shein, Temu, and Alibaba, has buoyed Meta’s ad revenue in recent quarters. Whether that trend will hold in the second quarter will be closely watched.

But behind Meta’s big headlines lurks a sobering reality: Its Reality Labs division continues to burn billions, and consensus growth expectations are reaching a price that demands flawless execution. In the first quarter of the year, that division posted $4.2 billion in losses — after posting $17.7 billion in losses in 2024; and that number could rise to nearly $20 billion in 2025, including roughly $4.9-$5.3 billion this quarter in some of the latest estimates. Much of Zuckerberg’s hardware and Metaverse vision has yet to pay off, and Meta’s virtual reality bets — including Horizon Worlds and Quest headsets — have struggled to reach mainstream users.

One bright spot: the Ray-Ban Meta smart glasses, co-developed with EssilorLuxottica. Zuckerberg claimed on the first-quarter earnings call that Meta’s glasses partnership has “tripled in sales” year-over-year, and there’s a plan to ramp production up to 10 million units annually. The glasses are now integrated with Meta’s standalone app and its Llama 4-powered AI assistant. Despite growing privacy concerns about the use of smart glasses, Meta has doubled down on the hardware with a $3.5 billion stake in EssilorLuxottica and a partnership with Oakley. But until scale materializes, the broader business must carry the burden of this long‑term bet.

Chatting up profits

Meta’s stealthiest monetization engine may not be in your feed — but in your chats. WhatsApp continues to be a slow-burn opportunity for Meta with analysts increasingly bullish on its revenue potential, and this quarter could be pivotal in showing early signs of that promise. Meta launched: ads in WhatsApp’s Updates tab, paid channel subscriptions, and promoted business channels. And the company has begun integrating AI agents into WhatsApp — including Llama 3-powered customer service agents, assistants for small businesses, and early-stage virtual shopping guides — while quietly expanding its business messaging platform globally. 

WhatsApp now boasts over 200 million businesses using its tools, according to Bank of America, and analysts believe its long-term revenue potential could exceed $20 billion annually if even a modest share of users adopt AI-powered services or business messaging. Click-to-message ad revenue already surpassed $10 billion a year — and could rise as high as $40 billion, per projections from Wolfe Research and William Blair’s Ralph Schackart, who estimates that WhatsApp could generate $5 billion in incremental profit by 2028. JPMorgan has flagged WhatsApp as a key growth driver over the next two years, and Q2 could offer early signs of that shift as business adoption accelerates. Still, regulatory pressure and privacy backlash — particularly in Europe and India — could cloud the upside. If there are any new details or color on engagement or monetization rates during Wednesday’s earnings call with Zuckerberg and other leadership, this area could be a headline-maker.

Threads is quietly scaling, too. But monetization on Threads remains in the early innings. Any color on time spent, advertiser tests, or user retention — especially in context with WhatsApp and Instagram monetization strategies — could influence analyst sentiment. Investors will also look for updates on time spent and daily active users across Facebook and Instagram, especially in North America and Europe, where growth has leveled off. Any slowdown in user engagement could pressure ad revenue and signal saturation in Meta’s core business. Reels continues to close the monetization gap with Stories, but investors will want fresh data on ad load and click-through rates — particularly as TikTok faces continued regulatory scrutiny and a potential ownership change in the U.S.

The price of being everywhere

On the Llama front, Meta’s open‑source models remain central to its identity, but benchmarks show that Llama 3 is still trailing GPT‑4o and Claude 3.5 in many areas, while Llama 4 (internally dubbed “Behemoth”) has seen delays. Meta has leaned heavily into distribution instead — embedding Llama into WhatsApp, Instagram, Messenger, and smart glasses — and hoping that broad usage can compensate for raw model rankings. Unlike OpenAI or Anthropic, Meta has yet to roll out a paid API or enterprise tier for Llama — a choice that reflects its open-source stance but also limits near-term monetization. Questions linger: Is Meta building models to lead… or just to distribute? Analysts will tune in for engagement and retention metrics as validation.

Margin discipline is another major line of scrutiny. Capex is ballooning, but Wall Street is especially sensitive right now to operating leverage. Meta’s margins slipped slightly last quarter (from 38% to 37%), and if they erode further — particularly on hiring, data center burn, or Reality Labs costs — valuation will be vulnerable. Still, with nearly $60 billion in cash and strong free cash flow, Meta’s capital allocation (including potential buybacks) may help cushion investor reaction if guidance disappoints.

And regulatory risks lurk in every corner of Meta’s AI agenda. The company has refused to sign the European Union’s voluntary AI Code of Practice, arguing that the guidelines introduce legal uncertainty beyond the existing AI Act, elevating tensions with Brussels as the legislation takes effect by August 2026. Courtrooms, too, are part of Meta’s battleground. The company is wrestling with regulatory crackdowns in Europe following an April 2025 €200 million Digital Markets Act fine, which forces it to offer limited, data-lite versions of its ad products in Europe, already slashing SME ad conversions by as much as 70%. Meta is appealing, but the ruling may force strategic changes before Q3. Elsewhere, Meta faces continued legal fights over content moderation, labor practices, data transfers, tax structures, and alleged misconduct.

Beyond the second quarter, Meta’s third-quarter guidance could set the tone for the stock. Investors will be sensitive to signs of softening ad budgets or rising FX headwinds, especially as seasonal ad spending patterns shift in a high-rate environment.

Ultimately, Wall Street will be watching to see whether Meta can thread the needle: turning sky-high infrastructure and hiring costs into meaningful product and revenue gains while fending off (and surpassing) rivals in the AI race. With Meta’s stock up over 19% year to date and currently trading near all-time highs, the stakes couldn’t be higher — and the margin for error couldn’t be smaller.