Waymo, the self-driving technology company owned by Alphabet, may soon introduce a new digital co-passenger inside its robotaxis. The company is reportedly experimenting with integrating Google’s Gemini AI chatbot into its autonomous vehicles, allowing riders to converse with an onboard assistant that can answer questions and manage limited in-car functions. If rolled out publicly, the feature could significantly change how passengers interact with driverless taxis.
The development surfaced after app researcher Jane Manchun Wong discovered references to the unreleased Gemini integration hidden within Waymo’s mobile app code. According to her findings, the system is governed by a detailed internal document called the Waymo Ride Assistant Meta-Prompt, which defines how the AI should behave, respond, and interact with riders. The intention appears to be creating a friendly yet unobtrusive AI companion that enhances the ride without distracting from safety or comfort.
Based on the uncovered system prompts, Waymo’s Gemini-powered assistant is designed to communicate in short, clear responses, typically limited to one to three sentences. The assistant avoids technical jargon and focuses on being polite, reassuring, and easy to understand. When activated via the in-car display, it can greet riders by name using pre-approved phrases and may adapt greetings based on how frequently a passenger uses Waymo’s service.
Beyond casual conversation, the AI assistant can reportedly control certain vehicle features such as cabin temperature, interior lighting, and music playback. However, its capabilities are intentionally restricted. Gemini cannot modify routes, explain real-time driving behavior, adjust windows, control volume levels, or interfere with navigation. When riders request unsupported features, the assistant is programmed to respond politely with optimistic phrases like stating that the feature is not yet available.
The AI is also capable of answering general questions unrelated to driving, such as weather updates, trivia, or local business hours. However, it will not make reservations, order food, or handle emergency situations, likely due to safety, privacy, and liability concerns. Special trigger phrases can also prompt the assistant to stop speaking, giving passengers full control over their interaction experience.
One of the most important design principles behind Waymo’s Gemini experiment is a strict separation between the conversational assistant and the autonomous driving system known as the Waymo Driver. If a passenger asks how the vehicle sees the road or makes driving decisions, Gemini is instructed to clarify that those tasks are handled by the Waymo Driver using sensors, cameras, and advanced software, rather than presenting itself as the entity in control.
The assistant is also prohibited from discussing accidents, speculating about incidents involving Waymo vehicles, or commenting on videos related to crashes. Internal guidelines reportedly instruct Gemini not to defend or apologize for the driving system’s performance and to avoid acting as a spokesperson. Mentions of competitors such as Tesla or the now-defunct Cruise are also carefully managed to prevent controversial or misleading discussions.
Waymo’s experiment comes at a time when automakers and mobility companies are racing to integrate artificial intelligence into vehicles. While some competitors are focusing on highly conversational AI companions, Waymo’s approach with Gemini appears more practical and travel-focused, prioritizing clarity, safety, and comfort over extended dialogue. If the integration becomes available to the public, Waymo’s robotaxis could soon offer more than just autonomous transportation. Passengers may experience a new kind of journey, one where a calm, intelligent digital assistant helps answer questions, manage comfort settings, and make driverless travel feel more personal and reassuring.
