Tesla FSD and GM Super Cruise Highlight Limits of Vehicle AI

Seminal self-driving technologies have blind spots, raising safety concerns as adoption grows.

Mar. 22, 2026 at 3:50pm

Got story updates? Submit your updates here. ›

Tesla's Full Self-Driving (FSD) and General Motors' Super Cruise are pioneering autonomous driving technologies, but they are not without their limitations. While these systems can greatly improve highway safety, they still have blind spots that can lead to dangerous situations if drivers don't fully understand their capabilities and shortcomings. Incidents like a Cadillac Lyriq failing to react to road construction and a Tesla Cybertruck continuing straight on an overpass curve highlight the need for better driver education around the current state of vehicle AI.

Why it matters

As more advanced driver assistance systems (ADAS) like FSD and Super Cruise become widespread, it's crucial that drivers understand their limitations to avoid overreliance and potential crashes. These technologies represent a major step toward fully autonomous vehicles, but the transition period where humans and AI share control of the driving task requires careful navigation.

The details

Tesla's FSD pushes the boundaries of ADAS by automating the entire driving experience, even on local roads, while GM's Super Cruise is limited to highway driving. Both systems have encountered issues, like the Cadillac Lyriq failing to react to road construction and a Tesla Cybertruck continuing straight on an overpass curve. Experts say these incidents underscore a systemic problem - drivers often make assumptions about AI capabilities that aren't accurate, leading to dangerous situations. Vehicle AI still struggles with 'unexpected' scenarios and 'long-tail' situations that human drivers handle with ease.

  • In March 2026, the author tested Tesla FSD and GM Super Cruise in various locations in Southern California.
  • A recent accident involving a Tesla Cybertruck on Autopilot occurred, where the vehicle failed to properly navigate a curve on an overpass.

The players

Tesla Full Self-Driving (FSD)

A pioneering autonomous driving technology developed by Tesla that can automate the entire driving experience, even on local roads.

GM Super Cruise

An advanced driver assistance system developed by General Motors that automates highway driving.

Jason Corso

Chief Scientist and Cofounder at Voxel51 and Toyota Professor of AI at the University of Michigan, who provided expert commentary on the limitations of current vehicle AI systems.

Got photos? Submit your photos here. ›

What they’re saying

“Modern ADAS and AV systems are already highly capable, but they are only as reliable as the data they've been trained on. While AV companies collect petabytes of driving data, scale alone does not guarantee comprehensive coverage. What matters is whether models are exposed to the right scenarios—especially rare edge cases (unusual actors, lighting conditions) and long-tail situations (human unpredictability, ambiguous signage, construction zones).”

— Jason Corso, Chief Scientist and Cofounder at Voxel51, Toyota Professor of AI at the University of Michigan

“Do I worry about people understanding the limits of ADAS technology? Much less than you'd think. Why? There is already strong evidence these systems drive better than most humans.”

— Jason Corso, Chief Scientist and Cofounder at Voxel51, Toyota Professor of AI at the University of Michigan

What’s next

As vehicle AI continues to advance, automakers and regulators will need to focus on better educating drivers about the capabilities and limitations of these systems to ensure safe adoption.

The takeaway

While Tesla FSD and GM Super Cruise represent significant progress in autonomous driving, they still have blind spots that can lead to dangerous situations if drivers don't fully understand their capabilities. Improving driver education and transparency around the current state of vehicle AI will be crucial as these technologies become more widespread.