Microsoft tells sales team to push own AI models over OpenAI, Google, Anthropic

Microsoft has instructed its sales team to promote its own AI models as superior to offerings from OpenAI, Alphabet's Google, and Anthropic. The directive came during a fiscal 2027 strategy session on Tuesday, with Executive Vice President Jay Parikh telling staff: "Everyone else is selling parts — we're selling the full end-to-end system." This messaging frames Microsoft's advantage not as a single AI component but as comprehensive integration across its product portfolio.

During the strategy session, Copilot chief Jacob Andreou made a direct competitive comparison, claiming that Anthropic's Claude was "slower, less accurate and did not have strong security integrations across Microsoft's Office apps," according to Bloomberg. The session emphasized that Microsoft's in-house AI models are more efficient and less expensive than competing offerings from OpenAI and Anthropic—a notable shift given that both rivals have historically supplied the AI models powering many of Microsoft's own products.

This strategic pivot comes after significant developments in Microsoft's relationship with its external AI partners. In April, Microsoft ended OpenAI's exclusivity through a new deal structure. More recently, the company moved to replace OpenAI and Anthropic models in Word and Excel applications as part of a cost-reduction effort. Microsoft CEO Satya Nadella has publicly articulated the rationale, warning that companies "pay for intelligence twice" when they rely on third-party AI and that using outside models risks exposing valuable business knowledge to external vendors.

The broader AI market is entering a price-competition phase. Meta CEO Mark Zuckerberg has signaled aggressive pricing plans, with JPMorgan analyst Doug Anmuth reporting on Monday that Meta's new Model API could be priced roughly 25% below OpenAI and Anthropic. Microsoft's strategy—combining cost efficiency with full vertical integration—reflects similar pressure to demonstrate value as AI becomes a commodity component within larger software platforms.

Microsoft's move to prioritize its own AI models reflects a fundamental shift in how the technology giant approaches its AI strategy. The company has long relied on external partners—OpenAI and Anthropic—to supply the AI engines powering products like Copilot. However, that relationship has been gradually unwinding: in April, Microsoft ended OpenAI's exclusivity agreement, and more recently the company began replacing third-party models in core Office applications to reduce spending. The fiscal 2027 strategy session represents a formal directive to accelerate this transition, signaling that Microsoft views in-house AI as both cheaper and more strategically aligned with its broader product ecosystem.

CEO Satya Nadella's framing—that third-party AI models create a "double payment" problem and expose business data to external vendors—articulates the business case behind this shift. By controlling the full AI stack, Microsoft positions itself to offer customers what Executive Vice President Jay Parikh described as the "full end-to-end system" rather than disconnected components. The competitive landscape is intensifying on price as well: Meta's announced pricing, roughly 25% below OpenAI and Anthropic, adds pressure on all players to demonstrate cost efficiency. Microsoft's strategy suggests it believes controlling its own models will ultimately deliver both better margins and stronger integration with its dominant Office suite.