Waymo Is Watching! The Robotaxi Giant’s Next Move Could Change Privacy Forever

Waymo, Alphabet Inc.’s autonomous driving subsidiary, is taking a bold new direction in the way it trains its artificial intelligence systems. According to a draft privacy policy uncovered by independent researcher Jane Manchun Wong, the company plans to use video data from interior cameras inside its robotaxis footage linked to passenger identities  to help train generative AI models.

This isn’t just a minor update to data policy. It marks a significant shift in how companies like Waymo treat the human data collected during everyday interactions. While self-driving cars already collect vast amounts of information from their surroundings road conditions, traffic flow, pedestrian movement turning the lens inward, toward the passengers, introduces a new set of ethical and privacy concerns.

What the Draft Policy Reveals

The unreleased version of Waymo’s privacy policy suggests that the company will not only use this interior data for improving its systems but may also share it for purposes like ad personalization. “Waymo may share data to improve and analyze its functionality and to tailor products, services, ads, and offers to your interests,” the draft policy says.

At face value, this sounds like standard corporate policy in the age of big data. But what sets it apart is the inclusion of camera footage from within the vehicle, especially when that footage is tied to a rider’s identity. In a space once considered semi-private — the backseat of a ride  passengers could now find themselves playing a silent role in training the AI models of tomorrow.

This development has raised questions from privacy advocates about how much transparency and control riders will have over their data, especially in such intimate settings.

An Opt-Out Option, But Will People Use It?

One piece of good news is that Waymo intends to roll out an opt-out feature, allowing riders to refuse the use of their personal data  including camera footage  for generative AI training. According to Waymo spokesperson Julia Ilina, the feature is currently in development and won’t require a change in the official privacy policy, but rather will serve as an extension of existing data preferences.

“Any data Waymo collects will adhere to the Waymo One Privacy Policy,” Ilina told TechCrunch. She further emphasized that the company will not share personal information with other Alphabet companies like Google or DeepMind, unless explicitly permitted by users or required to operate Waymo’s services.

Still, the effectiveness of this opt-out mechanism depends on how visible and accessible it is. As of now, the company has not finalized how it will inform users about this option — whether through an in-app notification, an email, or if users will need to find it buried within the app settings. Without a clear, proactive method of disclosure, many users may remain unaware of their choices.

Why Interior Cameras Raise Red Flags

While privacy policies from big tech firms have long included clauses about data use for service improvement and advertising, incorporating interior vehicle cameras introduces a unique dynamic. It’s one thing for an app to track your clicks or searches. It’s quite another to be recorded  in image and potentially audio  while you sit in a vehicle, possibly making phone calls, talking to others, or simply relaxing.

The draft policy reportedly allows riders to “opt out of Waymo, or its affiliates, using your personal information (including interior camera data associated with your identity) for training generative AI.” The inclusion of this specific language hints that the footage could be used to train models that understand more than just driving possibly even body language, voice tone, or emotional state.

That capability may help Waymo build safer, more responsive vehicles, but it also raises ethical concerns about consent, context, and surveillance.

What Will This Data Be Used For?

Ilina outlined several purposes for collecting this kind of data: improving safety, checking cleanliness, recovering lost items, enforcing in-car rules, assisting in emergencies, and enhancing the overall rider experience. These are arguably practical and even beneficial reasons.

However, the use of generative AI introduces broader implications. Interior footage could be used to model human behavior in ways that go beyond vehicle operations — perhaps to train virtual assistants, emotion-detection algorithms, or behavior prediction tools. This evolution could help make AI more responsive to human needs but would also make the data being collected far more sensitive.

Why Now? Waymo’s Business Pressures

Waymo’s push into new data strategies may also be motivated by financial concerns. While the company has shown massive growth now averaging over 200,000 paid robo taxi rides per week across major cities like Phoenix, San Francisco, Los Angeles, and Austin  it still hasn’t achieved profitability.

Alphabet doesn’t break out Waymo’s finances separately. Instead, it includes the company under its “Other Bets” category, which posted a $1.2 billion operating loss in 2024. Despite securing over $10 billion in funding from Alphabet and outside investors, Waymo remains a costly venture, with heavy spending on R&D, vehicle fleets, custom hardware, software, and infrastructure.

Expanding revenue sources  like personalized in-car advertising and AI-driven services may be necessary for Waymo to build a sustainable long-term business.

Ad Personalization, Opportunity or Invasion?

If Waymo begins using interior data to serve highly personalized ads, it may follow in the footsteps of other Alphabet companies like YouTube and Google Search. But bringing ads into the physical ride experience is different from digital targeting. Imagine being shown or hearing an ad based on your facial expression, posture, or conversation during a ride.

For some, that level of personalization may feel invasive, especially in a space traditionally viewed as neutral or private.

AI, Consent, and the Road Ahead

As AI systems grow more powerful, they need more human-centered data  real gestures, emotions, conversations  to learn. Waymo’s robotaxis, operating 24/7 in urban environments, offer a uniquely rich data source for these models.

But such opportunities come with responsibility. Companies like Waymo will need to go beyond offering basic opt-outs and embrace transparent consent mechanisms, regular audits, and clear disclosures to ensure that passengers are not unknowingly contributing to AI systems they may not agree with.

Some advocates suggest that true transparency means informed opt-in, not just opt-out. That means riders would need to explicitly agree to have their data used for AI training before it happens a higher bar that ensures ethical standards.

Navigating the Tightrope Between Innovation and Privacy

Waymo is at the cutting edge of autonomous driving and AI innovation. But as it moves to use increasingly personal data to fuel these advancements, it walks a tightrope between progress and privacy overreach.

While interior camera footage may enhance safety and user experience, its use for generative AI raises profound questions about autonomy, consent, and the future of personal data. If Waymo wants to maintain public trust while scaling its services, it will need to ensure its passengers understand  and agree with  how their data is being used.

Only time will tell whether Waymo can strike the right balance. For now, its journey toward full autonomy also requires navigating the equally complex road of digital ethics.

Writer: Chrycentia Henryana