Wall Street is responding positively to Microsoft’s strategic pivot in artificial intelligence, driven by a significant upgrade from a longtime skeptic. On Monday, Ben Reitzes of Melius Research changed his rating on the tech giant from Hold to Buy, reversing a stance he has maintained since downgrading the stock in February.
Reitzes, who had previously warned about the vulnerability of enterprise software to AI disruption, raised his price target to $665 from $465. This adjustment implies more than a 25% upside from Microsoft’s trading level of approximately $525. The upgrade coincides with a broader market optimism surrounding the company’s leadership in the next phase of AI adoption.
“The adults are in charge,” Reitzes wrote in his research note, arguing that Microsoft’s experienced management team will be rewarded by both clients and investors. He suggested that recent public concerns from AI industry figures, including Anthropic CEO Dario Amodei regarding AI safety, have inadvertently strengthened the demand for Microsoft’s cybersecurity offerings. These tools are increasingly seen as essential for protecting against data privacy breaches and rogue AI agents.
Reitzes noted that mature technology providers are likely to play a “key cog in the security, governance and the ‘AI CYA’ play” during these early innings of AI integration. He acknowledged that while he held similar bullish views last year, frustrations over Azure performance and Copilot issues led him to a more bearish position. However, recent product developments have shifted that perspective.
Key to the upgraded outlook is what Reitzes describes as the “repackaging of Copilot with Autopilot for agents,” which he believes enterprises will adopt to manage potential chaos. He argues that Microsoft’s prospects in core applications and Azure are stronger than previously thought, as the company appears more insulated from agentic threats.
“We see a credible path for Azure growth to exceed 50% as supply begins to catch up with demand, pricing gets firmer, and OpenAI ramps,” Reitzes stated. Although Microsoft remains a primary computing provider for OpenAI, the relationship has evolved since ChatGPT’s launch in late 2022. This evolution has allowed Microsoft to become more model-agnostic, focusing less on the AI model layer and more on the “harness layer.”
In this framework, the AI model acts as the brain, while the harness serves as the body capable of executing actions autonomously. Reitzes highlighted that as models become increasingly interchangeable and commoditized, Microsoft can capture value by owning the identity, governance, data, and billing infrastructure surrounding agentic workflows.
The analyst emphasized that fear of AI—whether from external hacks or the risk of tools going rogue—will solidify demand for platforms like Microsoft that offer robust protection. Enterprise customers appear willing to absorb additional costs for the security and governance provided by established software ecosystems, viewing them as a form of “AI insurance” for their boards.
Our board loves the governance narrative. It makes buying expensive enterprise AI packages much easier to justify.
25% upside sounds great on paper, but Azure supply constraints have plagued them for years. Hopeful, but cautious.
I just want Copilot to stop crashing. If they can fix that, the stock might actually deserve this target.
Finally, an analyst gets it. The harness layer is where the real money is, not just the models.
So now ‘AI insurance’ is the main selling point? Fear-mongering by design seems a bit cynical.