Instagram head Adam Mosseri has brushed aside longstanding claims that Meta secretly records users’ private conversations for ads, even as the company moves to deepen how it uses personal data to fuel its advertising system.
In a post on Wednesday, Mosseri described the notion that Meta activates phone microphones without consent as a “myth.” But privacy advocates argue the real concern is not audio surveillance — it’s how much data Meta is now poised to harvest from user interactions with artificial intelligence tools across Instagram, Facebook, and WhatsApp.
Beginning December 16, Meta’s updated privacy policy will allow the company to incorporate data from AI-powered features directly into its ad-targeting engine. That means conversations with chatbots, image-generation tools, or AI-assisted search could all become new signals for advertisers looking to profile users.
For years, users have been unsettled by ads that seem uncannily timed to private discussions. Mosseri insisted this is coincidence, pointing instead to the vast network of browsing behavior, location data, and advertiser inputs Meta already leverages. But analysts warn the expansion into AI-generated interactions could make targeting more intimate — and more difficult for users to understand or control.
“Meta doesn’t need to spy on your microphone,” one digital rights researcher said. “The sheer amount of data people feed into AI systems voluntarily is far richer and potentially more invasive.”
Meta, which has denied microphone surveillance since at least 2016, maintains its practices comply with data protection rules. Yet as AI becomes the company’s next big ad engine, critics say the line between useful personalization and exploitative profiling is only getting blurrier.
