Waymo plans to use in-cabin camera footage to train AI models
Waymo's unreleased privacy policy discovered by a security researcher indicates the company may use interior camera footage from its robotaxis to train generative AI models and tailor services to riders. Waymo disputed characterizations in the leaked draft and clarified the feature is under development with opt-out provisions, stating AI systems are not designed to identify individuals or serve targeted ads.
What happened
Security researcher Jane Manchun Wong discovered an unreleased Waymo privacy policy screenshot showing the company may use interior camera footage from its robotaxis to train generative AI models. According to the policy draft, Waymo plans to use this data to tailor products, services, ads, and offers to rider interests.
Waymo disputed some assertions in the leaked policy draft, stating the feature is still under development and contains placeholder text that does not accurately reflect its purpose. The company clarified that the feature will offer riders an opt-out option for data collection for machine learning training, and that its ML systems are not designed to identify individual people or use the data for targeted ads.
Reported impact
The discovery raised privacy concerns about surveillance of passengers in Waymo vehicles and potential use of intimate cabin footage for AI training purposes.
Index analysis
The leaked policy includes an opt-out provision for California riders to comply with the California Consumer Privacy Act. Wong noted that Waymo's approach offers riders the ability to opt out of having interior camera data associated with their identity used for generative AI training.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.