GM CEO Says AI Now Codes Self-Driving Cars, Shifting AV Strategy

What happened

GM CEO Mary Barra stated on the Q1 2026 earnings call that AI, not humans, is now coding self-driving cars, marking a strategic pivot after the company scrapped its Cruise robotaxi business and took a $6B EV writedown. The comment suggests GM is attempting to reduce AV development costs by leveraging AI, but it follows a pattern of large impairments and restructuring charges that have eroded GAAP profitability. The DeepValue report flags GM's core earnings as fragile, with North American truck/SUV margins still far from the 8-10% target and policy risks looming. Despite a 59% stock rally over 12 months to ~$85, the stock trades at 27x trailing EPS and 8.7x EV/EBITDA, leaving little margin of safety. The AI pivot may help contain AV spending, but it does not address the structural challenges of EV losses, China drag, and $110B net debt.

Implication

The shift to AI coding could accelerate GM's ADAS roadmap and reduce R&D burn, offering a potential positive catalyst for sentiment. However, the DeepValue thesis remains intact: GM's profitability is overly dependent on high-margin ICE trucks/SUVs, which face regulatory and cyclical headwinds, while EV and China restructuring charges persist. The stock's current price already embeds high expectations for sustained earnings and buybacks. Without concrete evidence of margin expansion to 8-10% and a halt to recurring impairments, the risk-reward is skewed to the downside. Investors should view this as a tactical cost-savings move rather than a transformative breakthrough, and trim positions on any rally toward the $90 trim level.

Thesis delta

The CEO's AI announcement alters the AV narrative from a failed robotaxi experiment to a potentially leaner autonomy effort, but it does not change the fundamental thesis that GM's core earnings are fragile and the stock overvalued. The market may temporarily interpret this as a positive catalyst, but the underlying risks—EV losses, China deterioration, and policy sensitivity—remain unaddressed. Without a demonstrable improvement in EBIT-adjusted margins and a reduction in 'one-time' charges, the bearish case for trimming above $90 persists.