Is Your Phone Listening to You? Why the Ads Feel So Specific (and What to Do About It)

The moment everyone recognises

You are sitting at the dinner table and you mention, almost in passing, that you should finally book that trip to Tasmania. Forty minutes later you open your phone and there is an ad for a Hobart hotel. Or you tell a friend you need new running shoes, and the next time you scroll, a sportswear ad is waiting for you. If that has ever happened, you have probably asked the same question. Is my phone listening to me?

It is a deeply uncomfortable feeling, and it is easy to understand why it sticks. The ad lands so close to the conversation that coincidence feels impossible. You start testing it. You say something unusual out loud, on purpose, and wait to see if an ad follows. Sometimes one does, and the feeling hardens into certainty.

Here is what this article will do. It will give you the honest short answer about whether your phone is recording you, explain why the ads still feel so specific even when the microphone is not involved, and walk you through the phone settings that genuinely reduce how much of your life is tracked. You do not need to be a privacy expert to follow along. The goal is modest: understand what is happening, and change two settings this week.

The short answer

The evidence says your phone is not secretly recording your conversations to sell you things. That is not me being naive about what companies do with data. It is that always-on eavesdropping would be a terrible way to build an advertising business, and no credible investigation has found one running at scale.

Think about what it would take. A phone maker or an app developer would need to record audio from millions of devices continuously, transcribe all of it, match the transcripts to advertising profiles, and do it all without the recording showing up in battery use, data use, or the permission screens that phones now display. Then it would need to survive the legal risk of recording private conversations without consent, which is illegal in most of Australia and in many other places. The whole plan is clumsy, expensive and dangerous, and the payoff is tiny, because there are far easier ways to learn what you might buy.

None of this means companies are not collecting a great deal of information about you. They are. It just means the microphone is not the main way they get it.

Why the ad lands so close anyway

So if your phone is not listening, why does the ad for the thing you only talked about show up an hour later? Because a surprising amount of your life is already visible to advertisers without a single word being spoken.

Start with location. Your phone knows where you are, and where you have been. Visit a camping shop on the weekend, and an ad for camping gear can follow you for weeks, even if you did not buy anything. Search is even more direct. You look up “best walking shoes for plantar fasciitis” on your phone, and suddenly every site you visit seems to know you are shopping for shoes.

Then there is the behaviour inside apps. The time you spend on a video, the things you pause to read, the way you browse a shopping site before closing it, all of it feeds a profile. Your phone also carries an advertising ID, a sort of anonymous label that lets different apps and websites recognise that the same person is moving between them. Put that together with cross-device tracking, where your phone, laptop and tablet are linked through a shared account, and a single profile starts to connect things you did across all of them.

Here is the part that explains the spooky feeling. A friend texts you a link to a product, or a colleague mentions a brand in a group chat, and that mention becomes searchable and linked to your devices. You forget you ever looked at the thing. All you remember is that you said it out loud, and then the ad arrived.

The coincidence problem

There is also a genuine trick of the mind at work, and it is worth naming because it explains so much of the unease.

Your brain is excellent at noticing when an ad matches something you recently said or thought. It is far less interested in the hundreds of ads you scrolled past that had nothing to do with your life. That asymmetry is called the frequency illusion, and it is the reason one spooky match feels like proof.

Imagine you see two hundred ads in a day. One of them, just one, is for a brand that came up in a conversation that morning. Which ad do you remember that night? The matching one. The other hundred and ninety-nine vanish. Your memory then tells a tidy story: I mentioned it, and the phone heard me. The alternative story, that the ad was a statistical near-miss out of a huge pool of guesses, is less dramatic and therefore less memorable.

This matters because it changes the question you should be asking. The real question is not “is the microphone on?” It is “how did they know I was in the market for this at all?” And the answer to that is the tracking described above, not eavesdropping.

Where the microphone genuinely is involved

To be fair to the worry, there are real places where a microphone is involved, and it is worth being precise about them.

Voice assistants such as Siri, Google Assistant and Alexa listen for their wake word, the “Hey Siri” or “OK Google” that switches them on. Only after that prompt do they record and send what you say, and they are supposed to stop once you finish the request. That is genuinely different from always-on listening for advertising, but it is also why people understandably get uneasy. A device that can hear a wake word is, in some sense, always hearing. Companies say the audio before the wake word stays on the device and is not used for ads, and most security researchers accept that, but the discomfort is reasonable.

There have also been real incidents involving smart speakers that made the news. A device that recorded a family’s conversation and sent it to a contact. Reports of assistants activating by mistake on words that sounded like the wake phrase. These were genuine privacy failures, and they are worth taking seriously. What they are not is evidence of a deliberate scheme to record conversations for advertising. If such a scheme existed, it would be a far bigger story than the occasional misfiring speaker.

Spyware is the other exception, and it is important not to confuse it with everyday advertising. Malware that records audio, tracks keystrokes and watches activity does exist, but it is a targeted threat. It is planted on a specific person’s device by someone who wants that particular person’s data, usually through a dodgy download or a phishing link. It is not how the ordinary ad you see on a Tuesday afternoon works. If you are worried about spyware specifically, that is a different conversation, and it starts with keeping your phone updated and being careful what you install.

What actually uses AI here

Artificial intelligence is genuinely part of this story, just not in the way the “phone is listening” theory imagines.

Ad platforms and data brokers use machine learning to make sense of the mountains of data described above. The systems look for patterns across millions of people: what kinds of content hold your attention, what you search for around the time you make a purchase, which ads you ignore. Over time they build a prediction of what you are likely to engage with next. That prediction is why an ad can feel personally written for you. It was not pulled from a recording of your voice. It was assembled from your behaviour, compared against the behaviour of everyone else, and aimed at you because the maths said you were likely to respond.

This is the same broad technology behind generative AI, the systems that can write text, answer questions and create images from a prompt. Both sides of AI, the predicting and the generating, have become part of how online content and advertising are made. If you are curious about the generative side and want a plain-English introduction, I have written a separate explainer on generative AI and how these intelligent systems work that covers it without the jargon.

It is worth sitting with that for a moment, because it changes where the unease belongs. A prediction engine fed on your behaviour can feel more intrusive than a microphone ever would, precisely because it does not need to hear you to know you. You can close the app, put the phone down and say nothing at all, and the profile still grows from what you tap, what you linger on and where you go. That is a real privacy trade-off, and it is the one worth arguing about. The microphone story is a distraction from it.

For this article, the takeaway is simpler. When an ad feels like it read your mind, what it actually did was read your habits. The AI involved is a prediction engine, not a listener.

What to do about it (a short, honest checklist)

There is a practical side to all of this, and it does not require throwing your phone in a river. The steps below genuinely reduce how much of your behaviour is tracked. They do not remove tracking completely, and I am not going to pretend otherwise. The goal is to take the creepiest parts down a few notches and to give you back a sense of control.

Audit your app permissions. Open your phone’s settings and look at which apps can use the microphone, camera and location. Most apps have no reason to use any of them. A weather app does not need your microphone. A flashlight app does not need your location. Switch off everything that is not obviously necessary, and you will notice no difference in how the apps work.

Turn off ad personalisation where it is offered. Both the major phone platforms and the big advertising systems let you limit how your data is used to tailor ads. The settings are buried in different places on different devices, but they exist, and switching them off stops the most targeted advertising.

Reset your advertising ID. This is the closest thing to a reset button for the anonymous label that ties your activity together. On both major phone systems you can delete or reset it, which breaks the link between your past behaviour and future ads. It is worth doing every few months.

Use the privacy checkups built into your phone. The main phone platforms now offer guided reviews of your privacy settings, walking you through location, tracking, advertising and data-sharing choices one screen at a time. They take a few minutes and they surface options most people never find on their own.

Keep your software updated. Updates patch security holes, and a phone that is up to date is a harder target for the spyware and malware described earlier. This is the least glamorous step on the list and one of the most important.

A note on expectations before you start. You may not notice the ads changing much at first, because the profiles that follow you are built up over years and take time to cool down. That does not mean the settings are pointless. It means you are doing the equivalent of locking a door that has been open for a long time. The benefit shows up gradually, and it shows up most in the things that stop following you from device to device.

One more habit fits here even though it is not a privacy setting. The same “stop and check before you act” reflex that protects you from a dodgy permission request is what keeps you safe from scam messages. Criminals impersonate the very brands whose ads fill your feeds, and a surprising number of people get caught because the message looks like it came from a company they trust. If you want the short version of how to spot those before you tap anything, I have covered it in a guide to spotting phishing emails and texts before you fall for them.

Ten minutes, twice a year

Here is the calm way to think about all of this.

The feeling that your phone is listening to you is completely understandable. It is the brain doing its job, noticing a coincidence and looking for a cause. But the tracking that creates those coincidences is real, and it is worth understanding and trimming. You do not need to become a privacy obsessive or read every terms-of-service document ever written. You need ten minutes, twice a year, to work through the settings that matter.

Set a reminder for the start of the year and the middle of the year. Open the privacy settings, run the checkup, reset the advertising ID, and glance at your app permissions to catch anything new that has sneaked in. That small routine does more for your peace of mind than any amount of worry, and it is honest about what it can and cannot do. Your phone will still know a fair amount about you. It just will not know quite so much, and more importantly, you will know exactly what it knows and why.

And the next time an ad appears right after you mention something out loud, you will have the full picture. It was not the microphone. It was a prediction built from your own behaviour, a coincidence your brain chose to notice, and a system that is very good at guessing. Now that you know how the trick works, it stops feeling like a trick at all.

Sources:

• Office of the Australian Information Commissioner (OAIC), Privacy and your personal information
• ACMA, Targeted advertising and your data
• Electronic Frontier Foundation, The myth of the hot mic