A proposed $68 million settlement was filed late Friday in federal court in San Jose, California, resolving a lawsuit that alleged Google’s voice-activated Assistant unlawfully recorded and shared private conversations. The deal remains subject to approval by U.S. District Judge Beth Labson Freeman.

The class-action lawsuit accused Google, part of Alphabet Inc., of monitoring users’ conversations after the Assistant was mistakenly activated by “hot words” like “Hey Google” or “Okay Google.” Plaintiffs argued the technology then captured private speech and used it to target users with advertisements.

The issue centers on so-called “false accepts,” when the Assistant misinterprets ordinary speech as a trigger phrase and begins recording. Users claimed these accidental activations led to intrusive data collection and unwanted ads, sparking privacy concerns similar to those raised in past disputes involving rival voice assistants.

Google Assistant operates similarly to Apple’s Siri, responding to voice commands without manual activation. Apple reached a comparable $95 million settlement in December 2024 over similar allegations.

Google has denied any wrongdoing but opted to settle to avoid the risks and costs of prolonged litigation, according to court filings. The company did not comment on the settlement on Monday.

If approved, the settlement would apply to consumers who purchased Google devices or experienced false accepts dating back to May 18, 2016.

Attorneys representing the plaintiffs may request up to one-third of the settlement—roughly $22.7 million—for legal fees, court documents show.