Smart speakers face scrutiny over radio listening claims

Smart speakers made by Apple, Amazon and Google are facing renewed scrutiny after accusations that the devices can listen to radio broadcasts in homes and use elements of that audio to refine advertising profiles and digital services, raising questions about consent, transparency and the limits of data collection inside private spaces.

The allegations centre on claims that smart speakers equipped with always-on microphones are capable of detecting audio signals from traditional FM or digital radio stations played nearby, analysing patterns such as station choice, programme type or advertising content, and using that information to enhance targeted advertising across connected platforms. While the companies involved deny any form of covert audio surveillance, privacy advocates argue that existing disclosures are vague and that users may not fully understand how ambient audio data can be processed.

Amazon, which markets Echo speakers powered by its Alexa voice assistant, says its devices only record and transmit audio after a recognised wake word is spoken, and that any data used to personalise services comes from explicit user interactions. Apple maintains that Siri-enabled devices are designed with privacy safeguards, including on-device processing and anonymisation, while Google states that its Nest and Assistant-enabled speakers do not monitor background media for advertising purposes.

Despite these assurances, concern has intensified as regulators and researchers examine how advances in machine learning make it easier to extract meaningful information from short or partial audio snippets. Audio fingerprinting technologies, long used to identify music tracks or measure broadcast reach, can recognise radio stations or adverts without recording full conversations. Privacy specialists say that, in theory, such techniques could allow companies to infer household habits without storing raw voice data.

The debate has been shaped by earlier controversies involving voice assistants. Several years ago, contractors working for technology firms disclosed that anonymised voice recordings were reviewed to improve speech recognition systems, prompting public backlash and changes to opt-in policies. Those incidents demonstrated that data flows within smart assistant ecosystems were broader than many consumers expected, fuelling suspicion about other potential uses of audio signals.

Regulatory attention has followed. Data protection authorities in Europe have repeatedly warned that consent for data collection must be specific and informed, particularly when devices operate in private homes. Consumer groups argue that even if smart speakers are not actively recording radio content, the possibility that metadata about ambient audio could be captured creates a chilling effect. They point to the difficulty users face in auditing or independently verifying what their devices are processing.

Industry representatives counter that the business incentives for covert eavesdropping are weak, given the reputational and legal risks. They also note that targeted advertising increasingly relies on declared preferences, app usage and account data rather than passive audio monitoring. Advertising effectiveness, they argue, can be improved without listening to background media, especially as platforms already integrate streaming services, search histories and smart home usage patterns.

Nevertheless, the claims have resonated with lawmakers examining the expansion of connected devices in everyday life. Smart speakers are now embedded in kitchens, bedrooms and living rooms, often linked to thermostats, lighting and televisions. This central role amplifies fears about function creep, where data collected for one purpose is later repurposed. Privacy advocates say stronger technical barriers, such as hardware-based microphone disconnects and clearer indicators of audio processing, are needed to rebuild trust.

Experts in digital ethics emphasise that the issue is not limited to whether radio listening occurs, but to broader questions about accountability. They argue that companies should publish detailed explanations of how audio data is handled, including whether any form of acoustic analysis happens when a device is idle. Transparency reports and independent audits, they say, would help dispel speculation and allow regulators to assess compliance.