Instagram head says Meta isn't eavesdropping, but AI is watching
Okay, so the head honcho at Instagram, Adam Mosseri, hopped on his account to address something that's been bugging people for ages: is Meta secretly listening to our conversations to target us with ads? It's an old conspiracy theory, and Meta's denied it before. But here's the kicker – they're about to use AI data to target ads, making the whole "listening" thing seem almost... unnecessary.
Think about it: have you ever just thought about something, and then bam, an ad for it pops up on your feed? It's creepy, right? Meta claims they don't need to eavesdrop to make these eerily accurate recommendations. Mosseri even calls the idea of recording conversations a "gross violation of privacy," though let's be real, Meta isn't exactly known for prioritizing user privacy above all else.
Back in 2016, Facebook (now Meta) flat-out said they don't use your phone's microphone for ad targeting. Zuckerberg himself testified about it before Congress. Now, Mosseri's reiterating this, saying you'd see a microphone light and your battery would drain if they were listening. Instead, he says their recommendation system works because advertisers share data about website visitors. They also show you ads based on what people with similar interests like. It's all algorithms, baby, and it's made Meta a ton of money.
Here's where it gets interesting: Meta's now using AI to make these ad decisions. If people thought they were being listened to before, just wait. Their new privacy policy lets them use data from your interactions with AI products. I think this is a potentially even more powerful signal than the old "people who like this also like that" system, since you're having personal conversations with AI chatbots.
Mosseri also brings up a good point: sometimes it's not even tech; it's just coincidence or how our brains work. You might've seen the ad before you had the conversation and not even realized it. We scroll so fast we internalize stuff without even knowing. Which makes me think... are we giving the machines too much credit here?
Source: TechCrunch