JetBrains Air launches to govern agentic software development

What happened

JetBrains has announced JetBrains Air, a new system of products designed to manage AI-powered software development workflows across individual developers, teams and organisations. Unveiled on 22 September, Air marks a shift in the company's focus from the individual developer workbench to the broader system through which agentic work is initiated, executed, coordinated, reviewed and governed.

The suite includes Air in JetBrains IDEs, which provides a complete agentic development experience for directing and orchestrating AI agents and verifying their output inside JetBrains' existing IDE products. It also includes Air Teams, aimed at coordinating and automating software-delivery workflows that combine human developers with autonomous agents, and Air Governance — a rebrand of the company's existing JetBrains Central product — which handles organisational policy, visibility and oversight of agentic activity.

JetBrains describes Air as an open, evolving system: some components are available now, while others will be introduced as the initiative develops. CEO Kirill Skrygan framed the announcement as a deliberate widening of scope after 26 years focused primarily on tooling for individual developers.

Why it matters

Air signals a structural shift in how software vendors think about AI agents in the development lifecycle: not as isolated coding assistants bolted onto an IDE, but as participants in a wider workflow that needs coordination, review and governance alongside human teams. By building products for orchestration and oversight rather than just code generation, JetBrains is acknowledging that agentic AI introduces new operational questions — who directs agents, how their work gets verified, and how organisations retain policy control as autonomous systems take on more delivery work.

For technology and transformation leaders, this points to agentic AI maturing from a productivity feature into infrastructure that requires its own management layer — much as cloud adoption eventually demanded its own governance and cost-control tooling. Organisations piloting AI-assisted development may need to think earlier about oversight and coordination, rather than treating governance as an afterthought once agent use scales.

The Renascence take

The interesting signal here isn't the IDE features — it's the governance layer. JetBrains is implicitly betting that the bottleneck in agentic development won't be model capability, but trust and coordination: knowing what an agent did, why, and whether it should be allowed to do it again.

Most organisations experimenting with AI agents are optimising for speed of output and skipping the harder design question: what does accountable delegation actually look like when the "employee" is autonomous software? A governance product arriving alongside the orchestration tools — rather than months later as a retrofit — is the more telling part of this announcement. Any operator introducing agentic workflows should treat visibility and review as a day-one requirement, not a compliance feature added after something goes wrong.