Waymo is reportedly planning to utilize data collected from its robotaxis, including footage from interior cameras linked to rider identities, to develop generative AI models. This information comes from an unreleased version of the company’s privacy policy discovered by researcher Jane Manchun Wong.

According to the draft language, Waymo may also share this data to create personalized advertisements, raising fresh concerns about the extent to which a rider’s behavior inside autonomous vehicles could be repurposed for AI training and marketing purposes.

The privacy page states that Waymo may share data to improve its functionality, analyze performance, and tailor products, services, ads, and offers to users’ interests. Users have the option to opt out of sharing their information with third parties, unless it is essential for the service to function.

While this language is standard in today’s digital landscape, the inclusion of cameras in the equation significantly raises the level of concern regarding user privacy.

Waymo provides riders with the option to prevent their personal information, as defined by California’s privacy laws, from being shared or sold. Additionally, riders can opt out of allowing Waymo or its affiliates to use their personal information, including interior camera data associated with their identity, for training generative AI models.

It remains unclear what specific interior data might be used to train generative AI models or what the intended applications of these models are. Furthermore, it is not evident what type of data the interior cameras capture, whether it includes facial expressions or body language, or whether Waymo is using this data to train in-house models or sharing it with other Alphabet companies working on AI, such as Google or DeepMind.

TechCrunch has reached out to Waymo for more information and will update this post if the company responds.

To date, Waymo is the only autonomous vehicle company generating revenue from robotaxi rides in the United States. As of February, the company is logging over 200,000 paid robotaxi rides every week via its commercial services in Los Angeles, San Francisco, Phoenix, and Austin. This represents a significant increase from 10,000 rides per week just two years ago and indicates potential for further growth as Waymo expands into new markets. The company plans to launch commercial services in Atlanta, Miami, and Washington D.C. over the next two years.

Despite these gains, Waymo is likely still incurring significant losses for Alphabet, which may explain why the company is exploring alternative revenue streams, such as in-vehicle advertising and data sharing for generative AI models.

Last year, Alphabet invested an additional $5 billion into Waymo, and the company raised a further $5.6 billion from external investors, increasing its valuation to over $45 billion.

Waymo continues to invest heavily in research and development, incurring significant costs associated with expansion, including growing its fleet, purchasing specialized equipment, maintaining vehicles, and developing charging infrastructure.

It is uncertain how close Waymo is to breaking even or achieving profitability. Alphabet does not provide a separate breakdown of Waymo’s financials in its earnings reports, instead including it in the “other bets” section of its balance sheet, which recorded an operating loss of $1.2 billion in 2024.