Measuring AI agent autonomy in practice

Source: Anthropic (McCain, Millar, Huang, ... Clark, Ganguli); Anthropic research report · 2026

AI agents are here, and already they’re being deployed across contexts that vary widely in consequence, from email triage to cyber espionage. Understanding this spectrum is critical for deploying AI safely, yet we know surprisingly little about how people actually use agents in the real world.

Claude Code is working autonomously for longer. Among the longest-running sessions, the length of time Claude Code works before stopping has nearly doubled in three months, from under 25 minutes to over 45 minutes. This increase is smooth across model releases, which suggests it isn’t purely a result of increased capabilities, and that existing models are capable of more autonomy than they exercise in practice.

Experienced users in Claude Code auto-approve more frequently, but interrupt more often. As users gain experience with Claude Code, they tend to stop reviewing each action and instead let Claude run autonomously, intervening only when needed. Among new users, roughly 20% of sessions use full auto-approve, which increases to over 40% as users gain experience.

Claude Code pauses for clarification more often than humans interrupt it. In addition to human-initiated stops, agent-initiated stops are also an important form of oversight in deployed systems. On the most complex tasks, Claude Code stops to ask for clarification more than twice as often as humans interrupt it.

Agents are used in risky domains, but not yet at scale. Most agent actions on our public API are low-risk and reversible. Software engineering accounted for nearly 50% of agentic activity, but we saw emerging usage in healthcare, finance, and cybersecurity.

Below, we present our methodology and findings in more detail, and end with recommendations for model developers, product developers, and policymakers. Our central conclusion is that effective oversight of agents will require new forms of post-deployment monitoring infrastructure and new human-AI interaction paradigms that help both the human and the AI manage autonomy and risk together.

Agents are difficult to study empirically. First, there is no agreed-upon definition of what an agent is. Second, agents are evolving quickly. Last year, many of the most sophisticated agents—including Claude Code—involved a single conversational thread, but today there are multi-agent systems that operate autonomously for hours. Finally, model providers have limited visibility into the architecture of their customers’ agents. For example, we have no reliable way to associate independent requests to our API into “sessions” of agentic activity. (We discuss this challenge in more detail at the end of this post.)

To start, for this study we adopted a definition of agents that is conceptually grounded and operationalizable: an agent is an AI system equipped with tools that allow it to take actions, like running code, calling external APIs, and sending messages to other agents.1 Studying the tools that agents use tells us a great deal about what they are doing in the world.

The more revealing signal is in the tail. The longest turns tell us the most about the most ambitious uses of Claude Code, and point to where autonomy is heading. Between October 2025 and January 2026, the 99.9th percentile turn duration nearly doubled, from under 25 minutes to over 45 minutes (Figure 1).

Notably, this increase is smooth across model releases. If autonomy were purely a function of model capability, we would expect sharp jumps with each new launch. The relative steadiness of this trend instead suggests several potential factors are at work, including power users building trust with the tool over time, applying Claude to increasingly ambitious tasks, and the product itself improving.

Both measurements point to a significant deployment overhang, where the autonomy models are capable of handling exceeds what they exercise in practice.

Neither capability evaluations nor our measurements alone give a complete picture of agent autonomy, but together they suggest that the latitude granted to models in practice lags behind what they can handle.

How do humans adapt how they work with agents over time? We found that people grant Claude Code more autonomy as they gain experience using it (Figure 2). Newer users (<50 sessions) employ full auto-approve roughly 20% of the time; by 750 sessions, this increases to over 40% of sessions.

Approving actions is only one method of supervising Claude Code. Users can also interrupt Claude while it is working to provide feedback. We find that interrupt rates increase with experience. New users (those with around 10 sessions) interrupt Claude in 5% of turns, while more experienced users interrupt in around 9% of turns (Figure 3).

Both interruptions and auto-approvals increase with experience. This apparent contradiction reflects a shift in users’ oversight strategy. New users are more likely to approve each action before it’s taken, and therefore rarely need to interrupt Claude mid-execution. Experienced users are more likely to let Claude work autonomously, stepping in when something goes wrong or needs redirection. The higher interrupt rate may also reflect active monitoring by users who have more honed instincts for when their intervention is needed. We expect the per-turn interrupt rate to eventually plateau as users settle into a stable oversight style, and indeed the curve may already be flattening among the most experienced users (though widening confidence intervals at higher session counts make this difficult to confirm).9We saw a similar pattern on our public API: 87% of tool calls on minimal-complexity tasks (like editing a line of code) have some form of human involvement, compared to only 67% of tool calls for high-complexity tasks (like autonomously finding zero-day exploits or writing a compiler).10 This may seem counterintuitive, but there are two likely explanations.