Seattle — In late September, approximately 50 executives from major Microsoft client organizations gathered at the Paccar Pavilion for a rare open forum with CEO Satya Nadella. The event highlighted the software giant’s current struggle to establish definitive leadership in the artificial intelligence sector, a challenge Nadella addressed by urging wider internal deployment of AI tools through a strategy he termed “diffusion inside the enterprise.”
“Like all things, change is always bottom up, top down, and middle up,” Nadella told the audience, emphasizing that transformation requires engagement at every organizational level. The 59-year-old CEO, who has spent more than half his career at the company, is attempting to replicate the success he achieved when he took over in 2014 and shifted Microsoft from legacy software toward cloud computing and subscriptions.
Despite boasting a market capitalization nearing $4 trillion, Microsoft has yet to emerge as the clear frontrunner in AI. NVIDIA controls the chip infrastructure, while Anthropic and OpenAI develop the leading language models. Meta recently launched a mass-market personal agent, and Google’s Gemini services have gained significant traction among both businesses and consumers. Although Microsoft was an early investor in OpenAI, that partnership has weakened, with OpenAI recently removing Microsoft’s exclusive cloud status and its license for OpenAI intellectual property becoming nonexclusive.
Wall Street remains cautious regarding Microsoft’s AI trajectory. As of Friday’s close, Microsoft stock had risen just 7% year-to-date, underperforming both its megacap peers and the broader Nasdaq index. It has not outpaced the S&P 500 in a calendar year since 2023. Tiffany Wade, co-portfolio manager at Columbia Threadneedle, noted that uncertainty surrounds Copilot, the flagship product of Nadella’s AI strategy. “I think there needs to be a little bit more maturity of the product before we figure out actually what the best solution or what pricing is,” Wade said.
Nadella’s previous tenure is viewed as a template for potential reinvention. After succeeding Steve Ballmer, he oversaw a dramatic turnaround that included abandoning perpetual Office licenses in favor of subscription-based software-as-a-service models and integrating rival platforms such as Oracle and Salesforce. He also orchestrated major acquisitions including LinkedIn, GitHub, and Minecraft developer Mojang.
Currently, Microsoft is attempting another structural shift toward consumption-based pricing, moving away from fixed subscriptions to charge clients based on usage intensity, particularly for AI tools. This model relies on “tokens,” the standard unit of measurement for AI computing consumption. Nadella stated during the July earnings call that this transition is essential to expanding the company’s total addressable market.
A key component of this strategy involves GitHub Copilot, which was relaunched in 2021 alongside OpenAI. While competitors like Cursor, owned by SpaceX, and Anthropic’s Claude Code have attracted developers with rapid innovation, Microsoft adjusted Copilot’s pricing to align with computing demand. The move appears to be paying off: the company reported 50 million users in July, up from 26 million the previous October. Kyle Daigle, GitHub’s operating chief, noted that the team had “learned lessons” and shared those insights across other Microsoft divisions eyeing similar pricing schemes.
Hayete Gallot, Microsoft’s executive vice president for security products, defended the consumption model, stating, “When customers consume, it means we add value.” Currently, Microsoft 365 subscribers pay an additional $30 per person monthly for Copilot, with 30 million seats sold against a commercial base exceeding 450 million users.
Microsoft’s AI ambitions also rest heavily on Mustafa Suleyman, the former DeepMind co-founder hired in 2024 to lead the company’s AI initiatives after its relationship with OpenAI cooled. Suleyman’s team has developed models for coding, image generation, and transcription, recently claiming their new speech-to-text technology is the most accurate globally. However, independent benchmarks suggest Microsoft’s proprietary models still trail those of Anthropic and OpenAI in raw intelligence.
Internal restructuring continues as Microsoft seeks to streamline its efforts. Earlier this year, the company merged its commercial and consumer Copilot engineering groups to boost sluggish adoption. Nadella has also publicly downplayed the necessity of building a standalone “frontier model,” arguing instead for a “frontier ecosystem” that integrates multiple AI providers. During a recent presentation, slides displayed Copilot’s ability to connect with models from Anthropic, OpenAI, SpaceX, and Microsoft AI itself.
As Microsoft prepares to host a Windows and Surface event in San Francisco focusing on local AI’s impact on personal computing, investors are watching to see if Nadella can once again steer the tech giant through a period of profound industry disruption.
Leave a Reply