Ever had a company promise you a gourmet meal and then serve you a single, sad cracker? That’s the vibe right now with Perplexity AI, and one user just got the boot for pointing out the menu fraud.

A paying Pro subscriber noticed the much-hyped “Deep Research” feature seemed to be taking a permanent nap. So, they did what any good detective would: they checked the company’s own official documentation and launch blog. Turns out, the evidence proved the feature was severely throttled, a shell of its advertised self. They presented this airtight case to the official subreddit, where it blew up with hundreds of upvotes and comments agreeing. The community verdict was in: something was fishy.

Here’s the hilarious part. Instead of a “Thanks for the feedback!” or even a corporate-speak “We’ll look into it,” the mods went full nuclear option. They permanently banned the user and deleted the entire popular thread. It’s the digital equivalent of a restaurant manager yelling “You’re banned!” after a customer reads the menu out loud to prove the steak is actually tofu. The real comedy is using a company’s own words as evidence and getting punished for it. It’s like getting grounded for quoting your mom’s house rules back to her.

This whole saga is a perfect internet culture moment. It combines the thrill of a detective story with the classic plot of “shoot the messenger.” The user wasn’t even trolling; they were doing unpaid quality assurance with cited sources! It’s a reminder that in the tech world, sometimes the most dangerous thing you can be is a paying customer who can read.

So, the next time a feature seems a little off, maybe think twice before citing the manual. You might just get a one-way ticket to Ban Town, population: you. The lesson is clear: you can deep research the problem, but you probably shouldn’t deep post about it.