Users of Microsoft’s Windows have grown frustrated with the company’s insistence on stuffing its Copilot AI chatbot into almost every corner of the widely-used operating system, earning it the pejorative nickname of “Microslop.”

That’s despite Microsoft admitting in its own Copilot terms of service that the AI shouldn’t be relied upon for virtually any important work.

“Copilot is for entertainment purposes only,” the lengthy document reads. “It can make mistakes, and it may not work as intended. Don’t rely on Copilot for important advice. Use Copilot at your own risk.”

It’s a bizarre self-contradiction, considering how steadfast Microsoft has been in its efforts to stuff Copilot into even simple Windows apps, like Microsoft Paint and the text editor Notepad, as well as productivity tools.

“Me personally, it’s not a good sign when a company won’t stand behind the accuracy of their product,” one Reddit user noted. “If Microsoft doesn’t trust Copilot, why should I?”

“1/3 of the entire American economy invested into a technology that’s for entertainment purposes only,” another user wrote. “Such confidence. I’m sure this will go well.”

“If a car came with a warning not to trust it and it has no specific purpose or design intent, you wouldn’t pay for it,” yet another argued.

A company spokesperson later clarified in a statement to PCMag that the odd phrasing is “legacy language from when Copilot originally launched as a search companion service in Bing.”

“As the product has evolved, that language is no longer reflective of how Copilot is used today and will be altered with our next update,” the spokesperson added.

Nonetheless, the eyebrow-raising language in its terms of service highlights a much broader trend, with AI companies touting the capabilities of their chatbots — while also evading responsibility any mistakes or made-up nonsense they may spit out.

Put simply, tech executives claim that the large language models behind tools like Copilot are the most important development since the Industrial Revolution. But they still have a strong tendency to hallucinate, making their outputs fundamentally unreliable.

Meanwhile, employees continue to be put under major pressure to make use of AI at all costs.

Microsoft’s competitors use similar language to cover for possible liabilities. For instance, Elon Musk’s xAI warns in its own terms of service that its chatbots may spit out hallucinations, “be offensive,” or “not accurately reflect real people, places or facts.”

The growing schism between the lofty promises of tech leaders and the sobering reality of what AI tools are capable of today remains a major point of contention as companies, including xAI (which was folded into SpaceX earlier this year), OpenAI, and Anthropic, gear up for potentially record-breaking IPOs.

And the cracks are already starting to show. Case in point, Amazon reported major outages earlier this year that were reportedly caused by faulty AI-generated code.

Managers ended up telling employees that more senior engineers will need to sign off any AI-assisted changes made by junior and mid-level engineers.

More on Microsoft: AI Is Killing Microsoft