They’re watching you while they sleep.

In case concerns over Meta glasses’ privacy-violating potential weren’t keeping people up at night. Customers are complaining that the tech titan’s state-of-the-art smart glasses are listening to them even while the device is set to sleep mode.

In three separate Reddit threads, the customers claimed that multiple iterations of the tech specs — which enable hands-free recording, communication and other functions — have recorded their conversations without being activated.

“I have observed the same behavior on THREE different pairs of Ray-Ban Meta glasses, including 2nd- and 3rd-generation models,” exclaimed one concerned wearer.

They claimed that following a recent firmware update, the devices started capturing their convos with a friend without being told “Hey Meta” — the magic word used to wake the accessories from their cybernetic slumber. The AI app reportedly even saved their “audio recordings and text transcripts of these conversations in my conversation history.”

“Meta AI does not respond to me, and I never intentionally activated it,” lamented the user. “I have noticed the white notification light activating while I am speaking normally.

The Redditor, who said they’re planning to file a class-action lawsuit if they can prove the issue isn’t isolated, asked if anyone else was being audio surveilled against their will.

As it turns out, they weren’t the only one that Meta allegedly had its AI … err, ear, on.

“I thought I was the only one so I reset the glasses and this was still happening,” said another owner. “I could see the LED light up in instances where I didn’t invoke Hey Meta.”

“Mine randomly activates the internal light now without any input from me,” insisted a third. “It does not make the wake-up noise voice command would the internal light just turns on. Something weird is going on!”

“I made a post about the AI turning on during random conversations. It’s not just you. I have gen 2,” one user insisted.

Meta reps have since addressed the issue in a statement, although they didn’t exactly diagnose what went wrong.

“This issue has been resolved,” a company spokesperson told The Post. “People can manage, review and delete voice logs of their interactions with Meta AI via settings at any time.”

Unfortunately, this is far from the first time the Meta glasses’ privacy concerns have come into focus.

Nightclubs and country governments alike have been banning the alleged spy-glasses over concerns that owners were using the tech to surveil people against their consent.

“I think the biggest problem with the glasses is that they’re very covert,” underground Brooklyn party promoter James, who has instituted such a policy, previously told The Post. “They’re like pervert glasses, [especially] if you don’t want to be observed and recorded when you’re partying.”

And while Meta has rolled out safeguards — including software that stops the glasses from recording if the indicator light is obscured — techsperts say that’s not enough.

“No single kill switch will be sufficient,” AI specialist Dr. Roman Yampolskiy previously told the Post.

The AI authority, who teaches computer science at the University of Louisville, said that companies like Meta need to install tamper-resistant recording lights, automatic face blurring and “meaningful penalties for abuse” if they really want to deter digital looky-loos.