Microsoft's AI strategy is in flux, according to a prominent technology analyst who says CEO Satya Nadella's recent pivot to a model-agnostic approach signals confusion about how the company will actually make money from artificial intelligence. Ben Reitzes, head of technology research at Melius Research, publicly criticized the shift during a Tuesday interview with CNBC, calling Microsoft's direction a "mess" and expressing reluctance to invest in hyperscaler companies until their AI monetization plans become clearer.

What Did Nadella Say About Microsoft's AI Direction?

During a recent interview with The Wall Street Journal, Nadella outlined a significant strategic shift for Microsoft. He criticized the dominance of frontier AI model builders like OpenAI and Anthropic, companies in which Microsoft itself has invested billions of dollars. Nadella stated that Microsoft would steer away from a future controlled entirely by these specialized AI firms, suggesting instead that the company would pursue a more balanced approach to artificial intelligence deployment.

"You can't say, hey, all white-collar jobs are gone and this could even be a weapon and we will use all the power to build data centers," Nadella said.

Satya Nadella, CEO at Microsoft

This statement reflects growing concerns within the tech industry about the sustainability of massive capital expenditures on AI infrastructure without clear paths to profitability. The comment also suggests Microsoft is reconsidering its heavy reliance on OpenAI's models as the foundation of its AI strategy.

Why Are Analysts Concerned About Microsoft's Consumption Model?

Reitzes's criticism centers on the ambiguity surrounding how Microsoft will charge customers for AI capabilities. The company is reportedly considering a hybrid approach combining partial consumption-based pricing with traditional licensing models. This uncertainty troubles investors who need clarity on revenue generation and profit margins.

"They're going to move to partial consumption, partial license. I mean, I don't know about you David, but call me when they figure it out," Reitzes said.

Ben Reitzes, Head of Technology Research at Melius Research

Reitzes emphasized that he would not recommend investing in hyperscaler companies like Microsoft until their AI monetization strategies become concrete. He noted that these firms are essentially transferring capital to semiconductor manufacturers without yet demonstrating how they will recoup those investments through AI services and products.

How Is This Strategy Affecting Market Sentiment?

Despite Reitzes's criticism, Microsoft shares were up nearly 2 percent on the day of his comments, suggesting the broader market has not yet fully absorbed concerns about the company's AI strategy. However, the semiconductor sector experienced significant pressure, with the Direxion Daily Semiconductor Bull 3X Shares exchange-traded fund (SOXL) falling more than 23 percent at the time of reporting.

The analyst's skepticism reflects a broader anxiety among investors about whether the massive capital investments in AI infrastructure will generate adequate returns. Reitzes pointed out that companies adopting AI will "absolutely" outperform those that do not, yet the question remains whether the infrastructure providers themselves will benefit proportionally from this trend.

What Are the Broader Implications for AI Investment?

Nadella's model-agnostic approach represents a departure from Microsoft's earlier strategy of betting heavily on exclusive partnerships with frontier AI model developers. By signaling openness to multiple AI models rather than reliance on a single provider, Microsoft is attempting to position itself as a neutral platform for AI deployment across enterprises. However, this strategy creates uncertainty about the company's competitive differentiation and pricing power.

The tension between Microsoft's massive infrastructure spending and unclear monetization reflects a fundamental challenge facing the entire AI industry. Companies are investing hundreds of billions of dollars in computing capacity and data centers, yet the path from infrastructure to profitable AI services remains murky. Reitzes suggested that investors should wait for greater clarity before committing capital to these companies.

How to Evaluate AI Company Investments During This Transition

• Monetization Clarity: Look for companies with explicit, detailed explanations of how they will charge customers for AI capabilities, including whether they use consumption-based pricing, subscription models, or hybrid approaches.
• Infrastructure Efficiency: Assess whether companies are generating measurable productivity gains and revenue growth from their AI investments, rather than simply spending capital on data centers without clear returns.
• Competitive Positioning: Evaluate whether companies are pursuing exclusive partnerships with specific AI model providers or adopting a model-agnostic approach that hedges against any single provider's dominance.
• Market Adoption Rates: Monitor how quickly enterprises are actually deploying AI tools and whether adoption is translating into meaningful revenue growth for infrastructure and software providers.

Reitzes did note that he is advising clients to "buy the dip" in semiconductor and technology stocks, suggesting that current market weakness may present buying opportunities for long-term investors. He compared the current volatility to previous market cycles, indicating that these types of selloffs are not unprecedented.

The debate between Nadella's strategic vision and Reitzes's skepticism highlights a critical moment in the AI industry's evolution. As artificial intelligence moves from experimental pilots to enterprise deployment, the companies providing infrastructure and platforms must demonstrate that their massive investments will generate sustainable, profitable returns. Until that clarity emerges, prominent analysts like Reitzes will likely remain cautious about recommending hyperscaler stocks to investors.