The latest earnings season has sharpened a new rule in the technology sector: investors are willing to tolerate unprecedented spending on artificial intelligence only when it clearly translates into accelerating growth.
The contrasting market reactions to Meta Platforms and Microsoft this week show how unforgiving that test has become.
Both companies slightly beat expectations for the December quarter.
Yet their forecasts told two very different stories about the payoff from AI, triggering sharply divergent share price moves.
Meta’s stock jumped about 10% in late trading, and was up by about 8% during Thursday’s premarket trading, while Microsoft’s fell 6.5%, underscoring how investors are now judging AI strategies not by ambition, but by measurable returns.
The shift reflects how the stakes have evolved since ChatGPT’s launch more than three years ago.
AI is no longer a futuristic promise; it is a near-term business metric.
Meta’s ad-driven AI versus Microsoft’s enterprise AI
Meta’s results highlighted the immediate commercial impact of artificial intelligence in its core business.
Revenue surged 24% in the December quarter, helped by AI-powered ad targeting that improved user engagement and advertising performance.
The company forecast first-quarter revenue growth of about 30% year on year, far ahead of market expectations.
“Meta’s headline numbers are a really interesting reflection of the market’s attitude toward spending in AI space,” said John Belton, portfolio manager at Gabelli Funds.
“All else equal, the market would typically be concerned, but they have a big revenue guide for the first quarter.”
The market’s response suggests that investors are more forgiving of heavy investment when the underlying business model is delivering visible gains.
Meta’s reliance on advertising — which accounts for about 98% of its revenue — has allowed it to quickly monetise AI improvements.
As Dan Salmon of New Street Research put it, “Jaw-dropping revenue acceleration trumps heavy investment, easily.”
Microsoft’s challenge, by contrast, lies in translating AI leadership into sustained financial momentum across a complex portfolio of businesses.
Its Azure cloud-computing division, widely seen as the key indicator of enterprise AI demand, grew 39% year on year in the December quarter — slightly above expectations but slower than the previous quarter.
“If you are bullish on this name, you think Azure can grow north of 40%,” said Jackson Ader, a software analyst at KeyBanc Capital, in an interview, WSJ reported.
“They didn’t, and the guidance makes it seem like that will be more difficult.”
Growth versus capital intensity
The divergence in investor sentiment is also rooted in the scale of spending.
Meta expects data-centre investments to surge as much as 87% this year to $135 billion and predicted a 43% jump in total expenses to $169 billion.
Yet its accelerating revenue trajectory has reassured markets that these costs may be manageable.
Microsoft, meanwhile, has been under pressure to justify its soaring capital expenditure after spending $37.5 billion in the October–December quarter.
Although the company expects capex to decline in the January–March period, concerns remain that its AI investments are outpacing the growth they are supposed to generate.
“Investors are disappointed that Microsoft’s capital expenditure spend and early foray into AI with ChatGPT is not significantly boosting earnings growth,” said Kathleen Brooks, research director at XTB.
She added that constrained data-centre capacity would further weigh on sentiment.
Microsoft has pointed to technical bottlenecks as one reason for slower cloud growth.
Chief Financial Officer Amy Hood said that if all newly available GPUs had been allocated to Azure rather than internal AI development, the growth rate would have exceeded 40%.
That explanation, however, has done little to calm markets increasingly focused on near-term returns rather than long-term potential.
OpenAI advantage or liability for Microsoft?
Another fault line lies in Microsoft’s deep relationship with OpenAI, which accounts for about 45% of its AI backlog.
While the partnership has cemented Microsoft’s leadership in enterprise AI, it has also raised concerns about concentration risk.
“Microsoft’s deep ties to OpenAI underpin its leadership in enterprise AI, but they also introduce concentration risk,” said Zavier Wong, market analyst at eToro.
The anxiety has been amplified by intensifying competition in the AI ecosystem.
OpenAI has reportedly issued internal warnings about rivals such as Google’s Gemini and Anthropic’s Claude, highlighting the fragility of first-mover advantages in a rapidly evolving market.
Meta faces fewer such dependencies.
Its AI strategy is largely built around internal systems designed to improve recommendation engines and advertising efficiency.
Chief Executive Mark Zuckerberg argued that AI would fundamentally transform user experiences and ad performance.
“Using AI will both improve the quality of the organic experience and of advertising,” he said, adding that superintelligence could create a “compounding effect” across Meta’s platforms.
Different business models, different investor tests
The contrasting reactions also reflect structural differences between the two companies.
Meta operates a relatively simple, consumer-focused business, where AI gains can be quickly monetised through advertising.
Microsoft’s portfolio spans enterprise software, cloud services, gaming, and consumer products, making the impact of AI harder to isolate and measure.
A sharp decline in Xbox sales, for instance, weighed on Microsoft’s More Personal Computing segment, diluting the perceived impact of AI-driven growth in Azure.
In this sense, the market’s verdict is not only about AI itself but about how directly AI can be tied to revenue growth.
The broader message from investors is becoming clear: ambition is no longer enough.
Companies that spend heavily on AI must demonstrate that the technology is already reshaping their financial trajectory.
The post Meta and Microsoft earnings highlight shifting investor mood on AI appeared first on Invezz