
General Motors will introduce its new OnStar AI Assistant in the 2027 Cadillac LYRIQ, bringing conversational voice controls, charging assistance and vehicle-specific answers to the electric SUV.
The feature will arrive through an over-the-air software update in the coming months, with additional GM vehicles to follow. Drivers will activate it by saying “Hey OnStar” before asking a question or requesting an adjustment.
Charging Help and Vehicle Controls
For EV drivers, one of the most useful capabilities could be checking whether the remaining battery charge is sufficient to reach a destination. The assistant can assess available range and help map charging stops.
It also supports everyday controls, including cabin temperature adjustments, dome lights, the glove box and charge port. Drivers can request charging updates, check tire pressure or ask what a dashboard warning means.
GM says the system combines live vehicle information with model-specific knowledge, allowing it to explain equipment such as one-pedal driving and lane keep assist. When a request requires human support, it can connect the driver with an OnStar Advisor.
New OnStar App Coming in 2027
The assistant forms part of a broader overhaul of GM’s connected services. A new OnStar app will begin replacing OnStar Guardian and the Chevrolet, Buick, GMC and Cadillac mobile apps in 2027.
Planned features include Connected Cameras, which can capture triggered incidents and send recordings and alerts to the app. OnStar Family will support destination sharing and pickup coordination, while connected safety guidance will draw on vehicle, weather and route information.
GM says customers will control feature activation and permissions. Camera recordings will be encrypted and time-limited, and the company says it cannot access Connected Camera recordings.
GM has not announced whether OnStar AI Assistant will require a subscription. Its usefulness for LYRIQ owners will depend on how reliably it handles everyday requests, alongside the eventual cost and availability across other models.





