Irish Regulators Could Investigate Google’s Voice Data Leak

Google Voice Data Leak Prompts Investigation From EU Watchdog

The news that Google hired contractors to listen to voice recordings from users has prompted a possible investigation by Irish data privacy regulators, according to a report by Bloomberg.

Belgian broadcaster VRT broke the news and Google responded in a blog post about it on Thursday (July 12). The Irish Data Protection Commission, which is Google’s main EU privacy watchdog, received a notification about the breach soon after. 

If the commission finds violations, it could fine Google heavily under EU data protection rules. Helen Dixon, Ireland’s commissioner for data protection, said there are about 20 ongoing probes into large technology companies like Apple, Instagram and Twitter. 

Google said it used the language experts with the sole intention of improving its voice assistant. They “review and transcribe a small set of queries to help us better understand those languages,” Google said in the post, adding that the transcription is “a critical part of the process of building speech technology, and is necessary to creating products like the Google Assistant.”

Google has chosen to go after the person that leaked the information.

“We just learned that one of these language reviewers has violated our data security policies by leaking confidential Dutch audio data,” Google said in the post. “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.”

Google said that only about 0.2 percent of audio snippets were reviewed, and that the snippets aren’t associated with Google accounts. Background noises or other conversations, it said, aren’t meant to be transcribed.

The person who leaked the information to a Belgian broadcaster said that they listened to upwards of 1,000 recordings and found that 153 were recorded accidentally and not meant to be heard.