Universities Are Telling Students AI Helps Them Learn. Some Educators Think That's the Lie Doing the Most Damage.

A researcher posted to Bluesky this week to clarify something after her AI rant got picked up by Times Higher Education[¹]: she isn't worried about students taking shortcuts. She's worried about students being lied to — by their universities, by ed-tech vendors, and by professors who've accepted the sales pitch without interrogating it. The post got 36 likes, which is modest by most measures, but the audience it reached — educators, researchers, academics watching the same slow normalization with growing unease — amplified it into something larger. It's the kind of post that travels because it says what a room full of people are already thinking but haven't quite put into words.

The concern isn't new, but the framing is sharpening. Where earlier critics of AI in education focused on cheating and academic dishonesty, the conversation among educators is shifting toward something harder to police: the idea that AI use isn't just bypassing learning but actively displacing it. Another Bluesky post put it plainly — that as AI becomes more integrated into classrooms and workflows, researchers are finding evidence that it erodes students' capacity for original thought and expression[²]. The word "eroding" is doing real load-bearing work there. It implies a gradual process, not a single transgression — something that happens to students who aren't even trying to cheat, just trying to keep up.

On the other side, there's a counter-frustration brewing that's distinctly less patient. A defiant Bluesky post[³] fired back at what it called "AI shaming" — dismissing the concern as performative moralizing from people who need to touch grass. The post drew a hard line: calling out AI use for cheating and environmental harm isn't shaming in any meaningful sense, it's a factual judgment. The vehemence of the rebuttal is its own kind of signal. When critics of AI criticism start sounding this irritated, it usually means the criticism has started landing somewhere uncomfortable. The AI ethics conversation around education is no longer a polite faculty-lounge debate — it has a combative, politically charged edge now, with "AI shaming" entering the vocabulary the way "cancel culture" did: as a move to delegitimize the concern rather than engage it.

What's conspicuously absent from the most-engaged posts this week is any serious institutional voice defending the optimistic case. The news coverage leans positive, and there are Bluesky accounts promoting Saudi Arabia's STEM initiatives and AI compute token programs for universities — but these read like press releases that wandered into the conversation uninvited. The posts generating actual engagement come from people describing a gap between what institutions are promising and what students are experiencing. That gap — between the ed-tech pitch and the classroom reality — is where the most charged energy lives right now.

The researcher whose post made it into Times Higher Education wasn't asking for AI to be banned. She was asking for honesty. That's a harder demand than a prohibition, because it requires institutions to admit that they've been selling something they don't fully understand yet. Whether universities have the appetite for that admission is the question the conversation is quietly circling — and based on what educators are saying publicly, most of them already know the answer.