Meta's glasses crackdown removed nine videos after a reporter sent the links

In mid-July, Instagram head Adam Mosseri said Meta did not want people "surreptitiously taking videos of other people and harassing them and then posting them on our platform," and that the company was trying to fight it "every way we can" ^[1]. In the month that followed, a Business Insider reporter found dozens of pickup-line and prank videos filmed on Meta glasses still live on popular and verified Instagram accounts, and Meta removed nine of them only after being sent a list of more than 20 links ^[2][3].

The gap between the two numbers is the story. Meta spokesperson Tracy Clayton told Business Insider that thousands of pieces of content had already been removed and several large accounts taken down, that the action falls under existing Community Standards on bullying and harassment, and that searches for "rizz" and "cold approach" are now blocked ^[4][5][6]. Aggregate takedown counts and blocked search strings are the cheapest things to produce. What a reporter with a browser found is the harder measure, and by that measure fewer than half of the specific items handed over were actioned ^[1].

One detail should worry anyone who signs off on a control. Creator Colin Allen, who describes himself as a lifestyle and comedy creator rather than a pickup artist, filmed the "secret code" clip that Meta removed ^[7][8]. He had used the same line on different women in several other videos, and those remain up ^[9]. That is enforcement resolving URLs, not patterns. A system that removes the clip a journalist names while leaving near-identical clips from the same account in place is not detecting behaviour; it is processing complaints.

The audit trail is worse. When Business Insider first reported Mosseri's comments, two large pickup artist accounts had been deactivated; days later one was back, which Meta said was an error, and it was re-banned after the reporter flagged it ^[10]. A ban that silently reverses itself and is only corrected by outside notice is a control with no verification loop.

Competitors landed in the same posture. TikTok, which says it has policies against harassing content, removed two of four videos sent to it, citing bullying and sexual harassment rules ^[11][12]. YouTube, which says it has policies covering certain harassment and dangerous prank videos, removed "several" videos from four channels sent to it and pulled one channel from monetization ^[13][14]. Across three platforms, the detection layer that worked was a journalist's inbox.

Carolina Are, a digital criminologist at the London School of Economics, told Business Insider that Meta tends to minimise its own responsibility by describing the problem as user-generated behaviour or an enforcement error, without recognising the systemic issues or what its policies and infrastructure enabled ^[15]. Meta's exposure here is unusual because it sells the camera as well as the feed: the videos are tagged as shot on "Meta glasses," some have millions of views, and the devices are already being called "creep glasses" by some people, against sales that have grown since the 2021 launch to top 7 million ^{[16][17][18][19]}.

Watch whether action moves upstream of press inquiries: account-level penalties for creators with repeat patterns rather than clip-level removals, and whether the "Meta glasses" tag itself becomes a detection signal instead of a marketing one. Watch also whether Meta explains how a banned account was reinstated. Until then, the operating lesson is narrow and portable: when a leader announces a policy, the number that matters is the detection rate, not the takedown count.