Regulators around the globe are monitoring what the big tech companies are up to and the recent decision by the US to proceed with antitrust investigations against big tech companies like Facebook, Google, Apple, and Amazon is one example of it.

Now Google has revealed that its human workers can listen to the recordings of Google Assistant users. Google has officially admitted that its Voice Assistant listens to human voice and understands it to share data with human voice experts. The findings were revealed after Google Assistant’s dutch language recordings were leaked. The Belgian broadcaster VRT has reviewed more than 1,000 audio clips and found out 153 of them have been captured accidentally. One of the leaked recordings includes a voice of woman’s son and baby grandchild.

Google said, “As part of our work to develop speech technology for more languages, we partner with language experts around the world who understand the nuances and accents of a specific language. These language experts review and transcribe a small set of queries to help us better understand those languages.”

The artificial intelligence (AI) technology listens to human voice, understand it and responds to queries. The AI-powered service is offered via the Assistant app on Android phones, or through the Google Home smart speakers. Assistant is only meant to send data to Google after the device detects a person interacting with it. Saying “Hello Google” triggers the Assistant to record the user’s voice.

On Thursday, Google acknowledged the breach after Belgian broadcaster VRT reviewed more than 1,000 audio clips and found 153 of them had incorrectly captured the voice. The recordings contained information like couple’s address and other sensitive information. Other parts of the recordings included phone calls, private conversations as well.

In the official blog post, Google’s product manager David Monsees said, “We just learned that one of these language reviewers has violated our data security policies by leaking confidential Dutch audio data. Our Security and Privacy Response teams have been activated on this issue, are investigating, and we will take action. We are conducting a full review of our safeguards in this space to prevent misconduct like this from happening again.”