AI-generated editorial illustration; not a documentary photograph.
JetBrains Research introduced the Artificial Intelligent Development Environments Framework, or AIDEs Framework, on September 18, 2026. Developed after nearly two years of product and market research, the framework classifies AI development tools by the work being delegated, the division of responsibility between people and systems, and the organizational setting in which development occurs.
Five Levels of Delegation
AIDEs defines five levels, beginning with L1 – Tool for narrow actions such as code completion. L2 – Assistant covers bounded multistep tasks, while L3 – General-purpose Executor shifts responsibility toward delivering a result but still requires expert task setting and review. JetBrains places current coding agents such as Claude Code and Codex roughly at L3.
The proposed L4 – Supervised Executor would manage an entire development function while escalating key decisions. L5 – Competence Center would receive strategic objectives and independently make implementation choices. JetBrains says real-world examples become scarce at L4 and does not present L5 as a description of current products.
Beyond Code Generation
The framework maps 35 software-creation activities across five groups, spanning work from initial concepts through delivery, maintenance and feedback. JetBrains argues that current AI adoption remains concentrated in brainstorming and coding, leaving much of the broader development lifecycle comparatively untouched.
Organization Changes the Constraints
A third dimension separates individual, SME and enterprise development. The categories account for differences in budgets, project scale, legacy systems, purchasing processes, security and compliance. This means the same AI system may receive different responsibilities depending on the environment and the principal's willingness to delegate.
Delegation, Not Benchmark Scores
AIDEs deliberately excludes model intelligence as a dimension. JetBrains maintains that benchmark performance does not directly determine practical autonomy: delegation depends on trust, context and the relationship between the principal assigning work and the agent performing it. The company also assumes humans will retain responsibility for task specification and oversight as delegation increases.
Source: The AIDEs Framework: How We Built a “Theory of Everything” for AI Development Tools
SIGNAL TITANS
Leave a Reply