Google has agreed to pay $68 million to settle a class-action lawsuit claiming its voice assistant recorded users without consent and leveraged that data for targeted advertising.
The preliminary settlement, filed last week in federal court in San Jose, follows allegations that Google Assistant activated and captured conversations even when users did not say ‘Hey Google’ or ‘OK Google.’
The lawsuit claims these so-called ‘false accepts’ resulted in the recording of sensitive personal conversations that were subsequently shared with advertisers.
Google has denied wrongdoing but opted to settle rather than face prolonged litigation.
While the two actions are not strictly related, the company is phasing out Google Assistant in favor of its Gemini AI platform, with the classic assistant set to disappear from most mobile devices by March 2026.
The latest news marks the second time in the past week that Google’s security credentials have been called into question, with researchers finding a flaw within Google Calendar that could allow bad actors to access private data.
When the Channel Becomes the Vulnerability
For enterprises leaning into voice-enabled customer service, the Google settlement exposes a fundamental tension in modern CX: customers want frictionless, voice-activated support, but they also want assurance that their devices are not listening when they should not be.
The lawsuit alleged that Google Assistant sometimes activated without any wake word at all, recording conversations that users believed were private.
According to court documents, plaintiffs said the system captured discussions about finances, work matters, and personal decisions. That information was then allegedly used to serve targeted ads, turning what should have been a secure channel into a data collection point.
The settlement applies to users who purchased Google devices or experienced false activations since May 18, 2016, with individual payouts depending on the number of claims filed, though consumers can submit claims for up to three Google devices.
This follows Apple’s $95 million settlement in December 2024 over similar allegations involving Siri.
Users are now receiving payouts ranging from about $8 to $40 per device in that case, which also centered on claims that the voice assistant recorded private conversations without user consent.
The CX Credibility Problem
The pattern emerging from these settlements should worry any contact center leader betting big on voice technology.
When customers cannot trust that their interactions are private and controlled, the entire value proposition of voice-enabled service starts to unravel.
The Google case is particularly troubling because it strikes at the core reliability claim of voice tech.
If a system is supposed to activate only when prompted but instead picks up background speech and treats it as a command, that is not just a technical glitch; it is a failure of the fundamental contract between the company and the customer.
In a nutshell, voice channels are only as strong as the trust customers have in them. If people start second-guessing whether their smart speaker or phone is listening when it should not be, they will either stop using those channels or approach them with suspicion.
What Contact Centers Should Be Thinking About
Although on the surface, the settlements against Google and Apple concern consumer devices, they also impact the broader ecosystem of voice-enabled technology, including the tools that power customer service interactions.
Contact centers using voice AI need to be asking hard questions about how their platforms handle voice data, when recording starts and stops, and what safeguards are in place to prevent unauthorized captures.
Transparency is the first line of defense. Customers should know exactly when they are being recorded, why, and what will happen to that data. That means clear disclosures at the start of interactions, not buried in terms of service that no one reads.
Technical safeguards matter just as much. If a system can be tricked into activating by background noise or misinterpreted speech, that is a design flaw that needs fixing.
Contact center leaders should be working with their tech partners to ensure that voice platforms have robust wake-word detection, tight permissions around data access, and audit trails that can prove when and why a recording occurred.
The Bigger Picture for Customer Trust
The Google settlement lands at a moment when customer trust in tech companies is already fragile.
Voice technology, with its always-on potential, sits at the intersection of those anxieties.
For CX leaders, this can quickly become a compliance and/or legal issue, as well as potentially impacting brand integrity.
If customers believe that your service channels are collecting data they did not agree to share, that perception will color every interaction they have with your company.
The damage is not just limited to the voice channel; it bleeds into trust across the entire customer relationship.
Google’s $68 million settlement is a reminder that the stakes are high, and for enterprises relying on voice-enabled CX, the implications are clear.
If customers cannot trust the channel, they will not use it, and all the AI and automation in the world will not fix that.
