General Motors plans to launch a proprietary in-vehicle artificial intelligence system later this year, a move that would put more of the carmaker’s software, vehicle knowledge and OnStar intelligence under one corporate roof. Anna Santos, GM director of product management of voice and AI/machine learning, said the new assistant will be more deeply integrated with GM vehicles than its recently launched Gemini AI assistant from Google.
Who Controls the Machine
Santos said, "Later this year, we'll be launching a more deeply integrated native AI assistant that combines conversational AI with GM vehicle knowledge and OnStar intelligence to create those capabilities that go beyond what a general purpose assistant can do." That’s the pitch: more integration, more data, more control. GM wants the assistant embedded deeper into the vehicle itself, not just sitting on top of it like a phone app with a nicer voice.
Santos said the new GM assistant, which she declined to name, will be able to better "understand the vehicle, the drive and our customers' needs, and make everyday ownership simpler." The language is polished, but the hierarchy is plain enough. GM is building a system that knows more about the car, the drive and the customer, while the customer gets told this is convenience.
GM last year announced the Gemini AI bot would launch this year in millions of 2022 model-year vehicles and newer, followed by a GM AI assist, but it did not provide additional details at the time. The rollout already shows the familiar corporate pattern: announce first, explain later, and let the machinery of ownership absorb the rest.
What the Driver Gets
With Gemini, customers can speak naturally without memorizing commands or repeating context. It is beginning to offer live sessions in which the bot will speak conversationally or play games such as trivia or 20 questions. It can also control some aspects of GM vehicles, such as temperature and radio controls, but generally operates as it would through a phone.
That’s the public-facing layer. Friendly, chatty, and wrapped in convenience. But the system still sits inside a corporate product, and GM decides what it can do, what data it can use and how far it reaches into the vehicle.
Santos said, "This is the beginning of a broader AI journey for us." She added, "There's a limit to what an AI that's just sort of sitting at the top level of the vehicle can do." The company’s own words make the direction clear. Top-level access isn’t enough. The apparatus wants deeper access.
GM is working with an unnamed large language model provider on technology to help with predictive maintenance, vehicle telemetry and other more auto-focused features. That could also include commands such as "kids setting" that would tailor music, seats, heating and cooling and door lock controls for children. The company is not just selling a voice assistant. It’s building a proprietary layer over the car’s functions, one that reaches into maintenance, telemetry and family settings.
Proprietary Data, Corporate Power
Santos said, "It's data that's going to be proprietary to GM, and our goal is to make sure that we're bringing the right technology forward to enable us to build the deep vehicle expertise that we want to be able to bring to the AI assistant." That sentence says the quiet part out loud. The data belongs to GM. The expertise belongs to GM. The assistant belongs to GM. The driver gets the interface, while the company gets the intelligence.
Separately, Space-Eyes, an Eric Trump-backed defense technology company, is set to go public in a $638 million SPAC deal. The company reflects rising investor interest in AI-enabled defense technologies such as drone detection and autonomous systems.
That’s the other side of the same machine. One company folds AI deeper into private vehicles. Another rides investor appetite for AI-enabled defense technologies, including drone detection and autonomous systems, through a $638 million SPAC deal. Different markets, same logic: more automation, more surveillance, more money for the people at the top while the rest are asked to call it progress.