General Motors’ Secret Weapon: 90% of Their Self-Driving Code Is Written by AI
General Motors just quietly revealed one of the most aggressive AI integrations in manufacturing history: nearly 90% of the code for its autonomous vehicle team is now generated by artificial intelligence. And AI is transforming how GM designs every vehicle, not just self-driving ones.
This isn’t a pilot program. This is a Fortune 500 automaker running its core engineering on AI.
What They Actually Did
GM’s AI transformation touches three layers of vehicle development.
Layer 1: Autonomous Vehicle Code. The company’s autonomy team – the group building GM’s self-driving capabilities – now generates roughly 90% of its code through AI. Human engineers review, direct, and refine, but the actual writing of code has shifted overwhelmingly to machine generation. This represents a fundamental change in how software-intensive vehicles are built.
Layer 2: Vehicle Design. Turning a concept vehicle sketch into a realistic animation used to take months of work by specialized teams. Now it takes days. AI handles the rendering, the visualization, and much of the iterative design work that once consumed entire departments.
Layer 3: Aerodynamic Testing. For decades, car companies tested aerodynamics by sculpting physical clay models. Multiple versions, each taking weeks to build and refine. GM has replaced much of this process with AI simulation – testing airflow patterns digitally before a single piece of clay is touched. The time and material savings are enormous.
GM raised its 2026 profit forecast after strong Q1 results, with AI-driven engineering efficiency cited as a contributing factor to faster development cycles and reduced R&D overhead.
Why This Actually Worked
GM’s approach reveals three principles that apply far beyond automotive:
They aimed AI at core work, not admin work. Most companies deploy AI on email summaries and meeting notes. GM deployed it on the thing they actually do for a living – designing and engineering vehicles. That’s where the real leverage lives.
They compressed iteration, not headcount. The goal wasn’t to eliminate designers or engineers. It was to let them iterate 10x faster. When you can test 50 aerodynamic profiles in a day instead of 3 in a month, the quality of the final product goes up dramatically.
They let AI do the building while humans do the judging. AI generates code, renders designs, and simulates physics. Humans decide what’s good, what’s not, and what direction to go next. This division of labor plays to each party’s strengths.
I’m Mike Partners. I founded AiExpert.org because I believe the strategies behind billion-dollar AI deployments should be accessible to every business owner. Here’s how to put this one into practice.
How to Apply This to Your Business
Even if you’re not building cars, GM’s strategy applies: Identify your longest creative or development cycle. Whatever your team spends the most time iterating on – product designs, architectural renders, marketing creative, software features – that’s your target. The longer the cycle, the more AI can compress it. Deploy AI tools that multiply iterations, not just speed. The real value isn’t doing the same thing faster. It’s doing 10 versions in the time it used to take to do one. More iterations means better final output. Look for AI tools specific to your creation process. Keep humans as editors, not executors. Let AI generate the first 80%. Have your best people refine the final 20%. This gets you machine speed with human judgment. It’s the same model GM uses, scaled to any size. GM proved that AI’s most powerful application isn’t replacing routine tasks – it’s supercharging the creative and engineering work at the heart of your business.
Frequently Asked Questions
How much of General Motors’ code is written by AI?
Nearly 90% of the code for GM’s autonomous vehicle team is now generated by AI. Human engineers review, direct, and refine the output, but actual code writing has shifted overwhelmingly to machine generation.
How is GM using AI in vehicle design?
GM uses AI to transform concept sketches into realistic animations in days instead of months. AI also handles aerodynamic testing through simulation, replacing much of the traditional clay model process. Mike Partners covers this at AiExpert.org.
What can small businesses learn from GM’s AI strategy?
Aim AI at your core creative or development work, not just admin tasks. Compress iteration cycles – test 50 variations instead of 3 – to dramatically improve final output quality regardless of industry.
Did GM’s AI adoption lead to layoffs?
GM compressed iteration cycles rather than headcount. The goal was letting designers and engineers iterate 10 times faster, improving product quality. GM raised its 2026 profit forecast partly due to AI-driven efficiency.
What AI tools can small businesses use for design work?
Use AI tools specific to your creation process to multiply iterations. Let AI generate the first 80% and have your best people refine the final 20%. This gets machine speed with human judgment at any scale.
