Meta Denies Microphone Spying, Confirms AI Chat Data Will Now Be Used for Targeted Ads
The head of Instagram, Adam Mosseri, has decided to publicly dispel a long-standing myth: that Meta secretly activates users’ microphones and eavesdrops on their conversations in order to deliver more precise advertising. This notion has circulated for years and grown into a popular conspiracy theory, despite the company’s repeated denials.
Mosseri’s statement, however, comes at a moment when Meta has announced plans to leverage data gathered from people’s interactions with its artificial intelligence–based products to refine ad targeting. Thus, while there was never a need to resort to recording conversations, the company is now poised to gain access to an even more powerful source of information.
He admitted that throughout his career at Instagram, he had frequently encountered this subject — including in personal conversations. Many users remain skeptical of how the company’s algorithms manage to deliver such precise recommendations. Even his own wife, Mosseri noted, once asked whether Meta was listening to their family’s phone calls.
For many, Meta’s services appear to “read their minds.” At times it seems as though simply thinking of a product is enough for it to appear in their feed. Meta has long insisted this is an illusion. As early as 2016, the company stated explicitly in its blog that smartphone microphones were not used for ad targeting or shaping the news feed. Two years later, Mark Zuckerberg reaffirmed this before the U.S. Congress, denying the use of voice data for advertising purposes.
Mosseri emphasized that users themselves would detect signs of covert recording — such as the microphone indicator lighting up or a rapid drain of battery power. He stressed that Meta’s recommendation system is so effective because of its close collaboration with advertisers: companies share data about visitors to their websites, which algorithms then analyze and cross-reference with the behavior of people with similar interests. This allows ads to be tailored to those most likely to respond.
Now Meta intends to expand this model by incorporating data gathered through interactions with its AI products, including the Meta AI chatbot. A new privacy policy, effective December 16, will grant the company the right to use such data in most countries. Since people often reveal personal interests, ideas, and plans in their conversations with chatbots, this provides algorithms with a uniquely valuable set of signals — richer than older mechanisms such as “people like you are also interested in this.”
Mosseri also observed that the uncanny alignment of ads with personal interests is not always technological. Sometimes it is mere coincidence or a psychological effect: a person may have glimpsed an advertisement before mentioning the product in conversation without consciously registering it, creating the impression that the platform is listening in.
Meta, then, once again denies surreptitious microphone use — but at the same time prepares to gather ever more data through AI tools, transforming users’ exchanges with virtual assistants into the next major source of advertising signals.
Support Our Threat Intelligence
If you find our technology report and cybersecurity news helpful, consider supporting our work.
Buy Me a Coffee
PayPal
