LaunchDarkly has launched AgentControl, a new product intended to give software teams real-time control over AI agents in production without redeploying the underlying application. The release extends the company’s runtime control platform into the growing market for AI operations and agent governance.
Because agent behaviour can change across models, prompts, and production contexts even when the surrounding application code stays the same, engineering teams are increasingly dealing with a different operational problem from conventional software releases. Configuration, quality control, governance, and intervention speed all become harder when live outputs are less predictable and spread across multiple teams and frameworks.
AgentControl is designed to address that by letting teams change agent behaviour at runtime, benchmark quality before updates reach live traffic, release changes through progressive exposure and guarded rollouts, and observe performance with trace-level visibility. LaunchDarkly said configuration changes propagate in under 200 milliseconds, allowing teams to reroute an agent, switch models, or trigger a fallback inside a live interaction.
Cameron Etezadi, chief technology officer of LaunchDarkly, said: “LaunchDarkly has always been about giving software teams control at runtime over what their software does in production. The hardest problems in AI, like model drift, unpredictable outputs, and the inability to intervene fast enough, turn out to be exactly the problems our platform was built to solve. We didn’t have to reinvent the platform; we just had to extend it to meet the demands of an AI SDLC and agentic-driven workflows.”
Cursor, which was referenced by LaunchDarkly as part of the announcement, framed that control layer as increasingly important as AI-assisted development moves deeper into enterprise production environments. Brian McCarthy, president of global revenue and field operations at Cursor, said: “Cursor is how the world’s leading enterprises are building with AI. As more AI-powered products and agentic capabilities reach production, runtime control becomes essential infrastructure alongside the development workflows and controls teams already trust.
“LaunchDarkly built an additional layer for that environment, and AgentControl extends it to the agent lifecycle in a way that complements how Cursor’s customers already build.”
LaunchDarkly’s move builds on the logic of feature flagging and release management, but applies it to software that can behave differently from one interaction to the next. That shift is becoming more relevant as companies move beyond pilots and start treating agents as part of core product and engineering infrastructure.
With more live AI systems requiring oversight after deployment, the value proposition here is straightforward: quicker intervention, fewer redeployments, and tighter operational control over software that does not always behave the same way twice.
You must be logged in to post a comment.