A casual conversation about a getaway, car rentals, and country houses , and then, a few hours later, Instagram is showing offers that are a perfect match. The suspicion arises on its own: "Are they listening to me?" That murmur has been circulating for years among dinner parties and online forums.
The parent company, Meta, has repeatedly denied the allegations, and now Instagram CEO Adam Mosseri is back with a video claiming there is no eavesdropping. This explanation comes just as Meta announces that conversations with its AI assistant will also influence ads and recommendations, a change that, for now, will not be implemented in the European Union.
Adam Mosseri denies the microphone myth on Instagram
Speaking directly, Mosseri maintains that Instagram does not activate the microphone to spy on what we say. According to him, doing so would be a serious privacy violation, as well as technically unfeasible due to battery consumption and the notifications that phone systems would display.
To explain why we see ads that seem to fit with recent conversations, he proposes four very common scenarios : perhaps we already searched for something related and forgot about it; maybe someone close to us interacted with that topic and the algorithm interpreted it as relevant; we may have seen that ad before without paying attention to it; or, simply, it is a coincidence.
“ We are not listening to you . We do not use the phone's microphone to spy. If we did, you would notice it from the battery and system indicators,” Mosseri essentially said in his public statement.
The executive admits that some people will remain skeptical, but insists that the platform is based on activity signals and affinities , not secret recordings. Ultimately, it's a combination of our digital footprint, the behavior of similar people, and statistics.
Technical evidence and public precedents
Meta has been publicly denying this for years: in 2016 the company dismissed these practices and in 2018 Mark Zuckerberg responded with a resounding “no” when the US Senate asked him if they used the microphone to target advertising.
Academia has also sought evidence. A team from Northeastern University analyzed thousands of Android apps —including those from Facebook—and found no evidence that they activated the microphone without the user's knowledge. They identified other methods of data collection, but no covert eavesdropping.
From a technical standpoint, both iOS and Android require explicit permission to access the microphone and display visible indicators when it's in use. Maintaining constant, undetectable listening would be difficult to conceal, as well as unsustainable due to daily energy consumption .
Given this scenario, the "always-on ear" hypothesis clashes with system safeguards , practical costs, and independent checks published to date.
Hyper-precise advertising, distrust, and the role of AI
The persistence of the myth is best understood in context: targeted advertising is so sophisticated that it's hard to believe there isn't a microphone behind it. This is compounded by Meta's history of privacy controversies, with scandals that have eroded the trust of some of its audience.
To add to the complexity, Meta has announced that, starting on certain dates, conversations with its AI assistant can be used to personalize ads and recommendations . The measure will not be rolled out in the European Union for the time being, but the timing of the announcement and Mosseri's clarification has not gone unnoticed.
Technically, it's not the same as "listening" to private conversations: here we're talking about voluntary interactions with an assistant within Meta's apps. Even so, for the average user, the line can be blurry and fuel a feeling of constant surveillance.
Credibility, therefore, cannot be restored with a video alone. Clear explanations, visible controls, and sustained transparency are needed for the advertising experience to cease being perceived as opaque or intrusive.
Between official denials, technical evidence and studies , and at the same time changes in personalization driven by AI , the debate will continue: today there is no proof of covert eavesdropping on Instagram, but the accuracy of the ads and the company's past continue to push many to distrust.
