For years, Microsoft’s Copilot tools have leaned almost entirely on OpenAI’s GPT models. That made sense—they were fast, capable, and already embedded deep into Microsoft’s stack. But last week, something interesting happened: Microsoft quietly added Anthropic’s Claude models into Microsoft 365 Copilot...
That’s not a small tweak. It’s a signal. Microsoft is no longer putting all its eggs in the GPT basket. And if you’ve been watching how Anthropic has grown over the past two years, you’ll know why this matters.
So, Who’s Anthropic and Why Claude?
Anthropic isn’t a household name (yet). The company was started in 2021 by ex-OpenAI researchers Dario and Daniela Amodei after philosophical differences on AI safety. Their pitch has been consistent: AI should be helpful, honest, and harmless. Sounds simple—but building models that actually stick to those principles is a huge challenge.
Their flagship model, Claude, has gained a reputation for two things: a steadier hand when it comes to safety, and surprising strength in reasoning. The latest versions—Claude Sonnet 4 and Claude Opus 4.1—are now competitive with GPT-4 on many benchmarks. More importantly, Anthropic has been pushing an open integration standard called the Model Context Protocol (MCP), making it easier for Claude to plug into business tools.
Why Microsoft Is Hedging Its Bets
OpenAI isn’t going anywhere—GPT models still lead in many areas like creative writing and brainstorming. But Microsoft knows one model doesn’t fit every use case. Anthropic gives them three things: diversification, a strong safety narrative, and competitive pressure on OpenAI.
Challenges remain—Anthropic hosts its own models, and Claude and GPT feel different in tone. But strategically? This is a smart move for Microsoft.
Claude vs. GPT: Not About Better, But Different
People keep asking, “Which is better—Claude or GPT?” That’s the wrong question. They’re different tools with different philosophies. Claude is steadier and more careful, GPT is flashier and more creative. I tested both: Claude excelled at summarizing compliance docs without hallucinations, GPT shined at rephrasing technical content into customer-friendly copy. Neither failed—they just excelled differently.
What This Means Going Forward
Choice is now the default in enterprise AI. With GPT and Claude side by side in Copilot, customers can see who’s faster, safer, or more affordable. Regulators are also watching, especially around copyright lawsuits. Microsoft may simply be spreading risk by working with both players.
Culturally, Anthropic’s “safety-first” approach and OpenAI’s “move fast” style will now co-exist under Microsoft’s roof. That’s fascinating to watch—and it signals the start of an era where multiple AI models coexist, each doing what it does best.
