MICROSOFT COACHES ITS OWN SALESPEOPLE TO TELL CUSTOMERS THAT OPENAI AND ANTHROPIC ARE SLOWER, LESS SECURE, AND NOT WORTH THE PRICE Microsoft held an internal strategy session on Tuesday where executives laid out a plan to turn its salesforce into something the company’s own partners would recognize: a machine for cutting OpenAI and Anthropic off at the knees. Executive Vice President Jay Parikh opened the session by telling the room that competitors were selling parts while Microsoft was selling the full system. Then executive Jacob Andreou went further, walking salespeople through a side-by-side comparison of Microsoft Copilot and Anthropic’s Claude, and telling them that Claude was slower, less accurate, and lacked the security integrations that enterprise customers actually required. The timing is not subtle. Microsoft quietly began replacing OpenAI and Anthropic models inside its flagship Word and Excel apps with its own in-house MAI family earlier this year, citing costs. What is new is that the company is now arming its salespeople to go on offense with those same arguments in front of customers. The irony is hard to miss. Microsoft invested roughly $13 billion in OpenAI and helped build Anthropic into what it is today. Now it is apparently training its own people to convince buyers that those models are the inferior choice. The AI partnership era is over. The AI vendor war has begun. Keywords: Microsoft Copilot, OpenAI, Anthropic, enterprise AI, MAI models