On Wednesday, Instagram chief Adam Mosseri took to his account to address and debunk the persistent rumor that the social media giant is covertly “listening” to users in order to serve them more targeted advertisements. The belief that Meta secretly activates users’ phone microphones to eavesdrop on conversations is a long-standing conspiracy theory—one that the company has denied multiple times.
Ironically, Mosseri’s attempt to clear up this misconception comes at a time when Meta has revealed plans to use data from users’ interactions with its AI tools to inform ad targeting across its suite of social apps.
Put simply, if Meta didn’t need to tap into your microphone to deliver uncannily precise suggestions before, it certainly won’t have to do so now.
Mosseri shared on Instagram that he’s had many discussions about whether Meta listens in on its users, noting that people are often amazed by how spot-on the company’s ad targeting can be. (He even mentioned that his wife has questioned him about it.)
By this point, most of us have either experienced or heard from someone who insists Meta must be secretly recording them to predict what they’ll click on. Sometimes, you might just be thinking about a subject or product, only to find it show up in your feed, making it feel as though Meta can read your mind.
Meta has consistently denied these allegations, explaining that it doesn’t need to record conversations to deliver such effective recommendations. (Mosseri also notes that this would be a “serious breach of privacy,” though Meta isn’t exactly known for putting user privacy first.)
Nevertheless, the company doesn’t have to literally “listen” in order to understand its users.
Back in 2016, when Meta was still called Facebook, the company published a blog post explicitly stating that it does not use your phone’s microphone to decide which ads or content to show you. Years later, CEO Mark Zuckerberg reiterated this stance before Congress, again denying that Meta collects audio data for these purposes.
With another privacy accusation it can refute, just as it prepares to collect even more user data, Mosseri repeats these assurances in his Instagram post.
He points out that if your phone’s microphone were active, you’d notice a light at the top of your screen and your battery would drain more quickly.
Instead, Mosseri clarifies that Meta’s recommendation engine is so effective because it collaborates with advertisers who provide details about visitors to their websites. This data enables Meta to display ads that are more relevant to users. Additionally, the platform shows ads based on the interests of people with similar profiles, using algorithms that have turned Meta into a highly profitable business.
Now, Meta plans to use AI to further refine its ad targeting. So, for those who already felt like they were being monitored, things may soon feel even more invasive. According to the company, its updated privacy policy—rolling out December 16—will permit Meta to use information from users’ interactions with its AI products in most regions as another data source. This could be even more influential than the “people who like this also like that” approach, since users are now sharing much more personal information with AI chatbots like Meta AI about their preferences, thoughts, and activities.
Mosseri also mentions that sometimes, the uncanny accuracy of recommendations isn’t due to technology alone—it can be chalked up to coincidence or psychological factors.
“It’s possible you saw that ad before your conversation and didn’t notice,” he explains. “We scroll through content quickly, including ads, and sometimes absorb things subconsciously, which can later influence what we talk about,” Mosseri adds.
