Public story · 2026-02-24 · source-backed

UC Berkeley L0-L5 Autonomy Framework

Issue

2026-02-24

Confidence

source-backed

Sources

Redaction

passed

Story

67-page framework defining six autonomy levels: L0 (no autonomy, direct human control) through L5 (full autonomy, users as observers). Addresses unsupervised execution, reward hacking, deceptive alignment, cascading compromises, and self-proliferation. Organized around NIST Govern/Map/Measure/Manage functions. Complements AWS Scoping Matrix (Scope 1-4). Most actionable governance framework for teams defining agent autonomy policies.

Source trail

67-page framework

Entities

Addresses Complements AWS Scoping Matrix L5 Autonomy Framework Manage Measure

UC Berkeley L0-L5 Autonomy Framework

Story

Related stories

Doubleword puts a date on open-weight parity: December 3, 2026

AWS Publishes Four-Scope Agent Security Framework

Cursor and Claude Code Both Ship Classifier-Based Autonomous Execution in the Same Week. The Industry Just Chose Its Architecture.

Source trail

Entities

SaaStr Hits 140% of Prior-Year Revenue with 1.25 Humans and 20+ AI Agents

Prismata Paper Reframes Web-Agent Prompt Injection As XSS

Lyzr closed a $100M Series B and let its own agent run the raise

Shopify's Farhan Thawar: "AI writes the code, AI reviews the code. Your job is to write the loops around it."

Stop managing agents by context-window length. Manage them by token count.

Provenance