The 5 levels of AI agent autonomy: learning from self-driving cars
October 3, 2025
How autonomous can AI agents become? We’re keen to access their power, but afraid of their unpredictability. Fortunately, it’s not the first time we’ve faced this problem. Self-driving cars have been grappling with autonomy for a while, and the industry developed a model to describe the different levels.
The autonomy spectrum runs from Level 1 (mostly manual, with some assistance) through Level 5 (full autonomy), giving car makers clear milestones while helping everyone understand where the industry stands.
This framework also works surprisingly well for AI agents, with “Level 0” representing the pre-AI era, and Level 5 being complete automation.
Source: The 5 levels of AI agent autonomy: learning from self-driving cars | AI Native Dev
Guy Podjani uses the framing of the levels of autonomy for self-driving vehicles for AI agents. Where are we at currently? Where might we end up?
I wasn’t alone 4-5 years ago in thinking that Level 5 autonomy for self-driving vehicles was at least within reach, but that has proven to be illusory.
Where might we end up in five years’ time? Those who are bullish suggest that fully autonomous agentic systems are just around the corner. The lesson I draw from self-driving vehicles is that might be further off than we think.
Related Speakers
