Microsoft is training its sales force to pitch Copilot against the very AI companies that have historically supplied the models running inside it. At an internal strategy meeting on Tuesday, executives briefed the sales organization on an FY27 plan to unfavorably compare products from OpenAI, Google, and Anthropic with Microsoft's own in-house systems, according to a Bloomberg report.
Executive Vice President Jay Parikh framed the pitch as a full-stack story. Executive Vice President Jacob Andreou reportedly took it a step further, presenting a side-by-side comparison of Copilot against Anthropic's Claude inside Microsoft's office applications. Andreou told the room that Claude was "slower and less accurate, and lacked the proper security integrations" when running against Word and Excel workflows, per Bloomberg's account.
The pitch itself — vendor trains sales team to knock competitors — is unremarkable. What is unusual is the target list. OpenAI is the company Microsoft has spent years funding, hosting, and reselling. Anthropic is the company Microsoft added to its Copilot stack as a second-source model provider. Both now sit on the same slide as the competition to beat.
“Everyone else is selling parts — we're selling the full end-to-end system. That's the story that we all need to get out there and tell in FY27.”— Jay Parikh, Microsoft Executive Vice President
Key facts
- 01At a Tuesday internal meeting, Microsoft executives outlined an FY27 sales strategy to negatively compare OpenAI, Google, and Anthropic products to its own.
- 02EVP Jacob Andreou reportedly called Anthropic's Claude 'slower and less accurate' than Copilot inside Microsoft's office apps, and lacking security integrations.
- 03Microsoft has been swapping OpenAI and Anthropic models out of Word and Excel in favor of its own models as a cost-cutting move.
- 04The Microsoft-OpenAI partnership was amended in April, dropping the exclusivity clause and clearing OpenAI to sell to Microsoft's competitors.
- 05EVP Jay Parikh framed the pitch as Microsoft 'selling the full end-to-end system' while rivals sell 'parts.'
The shift lines up with a partnership change from April, when Microsoft and OpenAI amended their agreement to drop the exclusivity clause that had bound OpenAI's API and models to Microsoft's cloud. OpenAI is now free to sell to Microsoft's direct competitors, and Microsoft is free to prioritize whichever model economics work best inside its own products. The revised deal turned a tightly coupled alliance into a more conventional supplier relationship.
Earlier this month, a separate report indicated Microsoft has been quietly swapping OpenAI and Anthropic models out of flagship apps like Word and Excel in favor of its own, framed as a cost-cutting decision. Andreou's Copilot-versus-Claude presentation reads as the sales-side justification for that engineering decision — if the internal models are cheaper and, per Microsoft, faster and better integrated, the sales team needs a story to tell enterprise buyers who ask why Copilot is no longer routed to a frontier lab's model.
Anthropic has not publicly responded to Andreou's characterization. Claude remains one of the top-ranked models on coding and long-context benchmarks in independent evaluations, and Anthropic has been aggressive about enterprise deployment, including the launch of Ode with Blackstone as a $1.5B AI implementation firm covered here earlier. Microsoft's internal framing — that Claude underperforms specifically inside Microsoft-owned surfaces like Word and Excel — is a narrower claim than a general model-quality argument, and it is a claim only Microsoft is currently in a position to test.
The financial context matters. Microsoft has faced pressure over the scale of its AI capex, with investors questioning whether the buildout will pay back inside a reasonable horizon. A sales narrative that positions Microsoft as the end-to-end provider — silicon, data center, model, application, distribution — rather than a reseller of someone else's model is a cleaner story to tell Wall Street. It also justifies vertical integration spending that a pure-reseller model would not.
The counterweight: Microsoft's own frontier-model track record inside consumer-facing chat has lagged OpenAI and Anthropic on public benchmarks, and enterprise buyers tend to notice when the underlying model changes. If Copilot's model swap produces a measurable quality drop in customer workflows, the sales pitch about security integrations and end-to-end control will not hold up against user complaints. Microsoft is betting that in-app integration and cost matter more to enterprise IT than raw model leaderboard position.
The Microsoft-OpenAI relationship is now formally competitive in the segments where both companies sell to enterprises, and Microsoft's sales pitch is the clearest signal yet of how that competition will be waged. For Anthropic and OpenAI, the read is that Microsoft's channel — one of the largest enterprise software sales organizations on earth — is now actively pointed at them inside every renewal conversation. That changes the go-to-market math for both labs and raises the value of every non-Microsoft distribution channel they can build.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
