"We’re updating the wording," Microsoft says after Copilot backlash

Microsoft is now pushing back on claims that Copilot is meant only for "entertainment," after its own terms of use caused a lot of confusion online (read more about that here).

The issue started when users spotted a section in Microsoft’s Copilot documentation stating the AI "may make mistakes," shouldn’t be trusted for important advice, and is intended for entertainment purposes only. The wording is clearly strange because it clashes with how Microsoft currently advertises Copilot - as a "productivity tool" across Windows 11 and Microsoft 365. Also, the integration of Copilot into the OS has been met with a lot of polarizing reception as of late.

In a statement shared with Windows Latest, Microsoft said that language is no longer accurate. The company explained that the "entertainment" part of the documentation goes back to the early Bing Chat days, when generative AI tools were still being used more carefully, and with restraint. According to Microsoft, Copilot has since "evolved", and the documentation will be updated to better reflect its current role in the OS and elsewhere.

The issue, however, isn’t just the wording itself but the disconnect between marketing and legal framing. While Microsoft promotes Copilot as a tool to get work done, its current disclaimers still include uncertainty, lack of guarantees, and user responsibility. Even outside the "entertainment" line in the document, Microsoft’s terms regarding "outputs may be inaccurate" and "users are responsible for any consequences" haven't changed - but as the source adds, that case holds for other AI tools as well. So while the company says Copilot is built for real work, the fine print just hasn’t caught up yet, that's all.