Is My Phone Listening to Me? UK Guide

You mention a new hip replacement to a friend over coffee, and an hour later an advert for orthopaedic cushions turns up on your phone. It feels like a coincidence too strange to be one. Plenty of people over 60 have had that exact moment and come away convinced their phone was eavesdropping.

It's a reasonable thing to worry about, and it deserves a straight answer rather than a sales pitch.

Is My Phone Really Listening to Me? The Short Answer

No, not the way it feels. Voice assistants such as Siri, Google Assistant and Alexa listen for a specific wake word, "Hey Siri", "OK Google", or "Alexa", and only send audio anywhere once that word is spoken. Outside of that trigger, the microphone is not actively transmitting anything to a company's servers.

The adverts that feel targeted almost always come from something else entirely: your search history, the apps you have installed, the websites you visit, and even your location. Companies build a detailed picture of your interests from that data, and their predictions are good enough to feel like mind-reading. AI chatbots such as ChatGPT are a separate matter again. Unless you specifically switch on a voice feature, they respond only to what you type, and have no access to your microphone at all.

Why It Feels Like Your Phone Is Listening

Modern advertising systems are built to spot patterns you would not notice yourself. If you searched for "orthopaedic cushion" on a tablet last month, browsed a mobility aids website, or even just walked past a pharmacy that sells them, an algorithmA step-by-step set of rules that tells a computer what to do. can connect those dots and serve you a relevant advert days later.

Part of what makes this feel so uncanny is that the tracking follows you across devices. If you are signed into the same Google or Apple account on your phone, tablet and laptop, a search on one device can shape the adverts you see on another. Add in loyalty cards, online shopping accounts and the websites that share data with advertising networks, and a company can build a surprisingly detailed picture of your interests without ever needing your microphone. A conversation about a hip replacement over coffee is easy to blame, but the more likely trigger is something you typed, clicked or browsed in the days beforehand.

Researchers who have studied this repeatedly find no evidence of phones secretly recording conversations for advertising. What they do find is that data-driven prediction has become so accurate that the difference is hard to spot from the outside. It is unsettling either way, but the fix is different depending on which one is actually happening: turning off your microphone stops eavesdropping, but it does nothing for tracking based on your browsing and search history.

What Voice Assistants Like Siri and Alexa Actually Do

Voice assistants use a small piece of hardware on the device itself to listen for one thing only, the wake word. That detection happens locally, without sending anything anywhere. Once the wake word is heard, the device starts recording and sends that clip to the company's servers to work out what you asked for.

Both Apple and Google have published clear statements confirming that Siri and Google Assistant do not record or transmit audio without that trigger. Apple's own privacy page confirms that no voice recordings are sent to its servers until the device hears "Hey Siri", and Google's guide to how Assistant is designed for privacy says the same for "Hey Google": nothing is sent or saved while the device is in standby. Mistakes do happen occasionally, background noise can sound close enough to a wake word to set it off, which is why it is worth reviewing your voice history now and then and deleting anything you would rather not keep. If you have an Amazon Echo at home, our guide covers whether it's safe to have Alexa in your home in more detail.

Do AI Chatbots Like ChatGPT Listen to Me?

This is a different question from voice assistants, and it is worth separating the two clearly. AI chat tools such as ChatGPT respond to what you type into a text box. There is no microphone involved unless you deliberately switch on a voice mode within the app, and when you do, the screen makes it obvious that it is listening.

The real privacy question with chatbots is not eavesdropping, it is what you choose to type. If you paste in personal details, medical information, or financial numbers, that text is stored by the company running the chatbot, and in some cases it can be used to improve the underlying AI model unless you turn that setting off. For a full rundown of what to keep out of a conversation, see our guide on what you should never tell ChatGPT.

If you do try a voice feature in a chatbot app, look for a microphone icon that changes colour or a waveform animation while it is active. That visual signal is deliberate: it is there so you always know when the microphone is genuinely switched on, rather than having to guess.

Your Right to Know: What UK Law Says

UK data protection law gives you more control here than most people realise. Under UK GDPR, you have the right to ask any company what personal data they hold about you and to request that it be deleted. This applies to tech companies exactly as it applies to your bank or your GP surgery.

The Information Commissioner's Office has a plain-English guide to protecting your privacy online, and a separate page setting out your right to have your data deleted. In practice, this means you can email or write to a company asking what data it holds on you, known as a subject access request, and it must respond within a month. If you would rather clear what a chatbot has stored about your conversations, our guide to your right to delete what a company holds on you walks through the steps.

Five Minutes to Check Your Phone's Microphone Settings

We checked these settings on both an Android phone and an iPhone, and the whole process took under five minutes on each.

• Open Settings, then Privacy (iPhone) or Apps and Permissions (Android), and look at Microphone
• Switch off microphone access for any app that has no obvious reason to need it
• Review your ad settings in your Google or Apple account and turn off ad personalisation if you would rather not see it
• Turn off voice assistants entirely if you never use them
• Check the privacy label shown in the app store before installing anything new

None of these steps require any technical skill, just a few taps. The ad-settings step is worth a closer look: on an iPhone it is Settings, then Privacy & Security, then Apple Advertising, where a single switch turns off personalised ads. On Android it is Settings, then Google, then Ads, where "Opt out of Ads Personalisation" does the same job. We tested both toggles while writing this guide, and in each case it took under a minute to find and switch off.

When It's Not Your Imagination: Real Warning Signs

Genuine microphone spying does happen, but it is caused by malware, not by ordinary apps or voice assistants working as intended. The National Cyber Security Centre's guidance on recovering a hacked device lists the genuine warning signs to watch for: a phone running much slower than usual, restarting by itself, apps closing unexpectedly or opening on their own, or pop-up boxes from programs you do not recognise asking you to do something unexpected. A battery that drains far faster than normal, a phone that runs hot without an obvious reason, or a sudden jump in your data usage are also worth checking, since they often point to something running quietly in the background. If several of these turn up together, it is worth running a security check or asking someone you trust to look at the phone with you.

Myth vs Reality

• Myth: AI chatbots secretly listen through your microphone. Reality: text chatbots only see what you type
• Myth: Your phone records everything you say. Reality: voice assistants only activate after a wake word
• Myth: There is nothing you can do about it. Reality: UK law gives you the right to see and delete what companies hold on you

Which?'s own research into how companies track you for advertising is a useful independent source if you want to understand more about what is being collected and why, and it is worth reading precisely because it has nothing to sell you.

If AI in general still makes you uneasy, our guide to common worries about AI covers the wider picture, and our AI safety guide is the natural next stop for a settings walkthrough that goes beyond just the microphone. The Academy course covers privacy settings step by step if you would like a guided version of all this.

FAQ