Alphabet’s self-driving technology subsidiary, Waymo, is gearing up to train its proprietary generative AI model using data collected from interior cameras in its robotaxis.
On Tuesday, TechCrunch reported that IT security researcher Jane Manchun Wong uncovered an unpublished privacy policy from Waymo. The document reveals that Waymo may use footage from in-vehicle cameras for AI training while granting passengers the right to opt out of this data collection.
Waymo had already collected personal data for service improvement and advertising under its existing privacy policy. However, the new feature would allow passengers to decline the sale, sharing, or use of their personal information for AI training.
Waymo spokesperson Julia Ilina clarified that this feature doesn’t alter the company’s privacy policy. “It simply offers passengers the choice to opt out of data collection for AI training purposes,” she added.
The company intends to utilize internal camera data for various purposes, including enhancing vehicle safety, monitoring cleanliness, tracking lost items, providing emergency assistance, and ensuring compliance with vehicle regulations. Waymo’s privacy policy stipulates that the collected data will not be shared with other Alphabet companies and will be used solely for Waymo’s service provision.
Currently, Waymo is the sole operator of robotaxi services in the United States, facilitating over 200,000 paid rides weekly across Los Angeles, San Francisco, Phoenix, and Austin. Waymo plans to expand its operations to Atlanta, Miami, and Washington D.C shortly.
Alphabet invested an additional $5 billion in Waymo last year, significantly boosting its operations. The company also secured $5.6 billion from external investors, pushing its valuation beyond $45 billion.
