Realtime AI News
WSO2 ships Agent Manager to rein in enterprise AI agent sprawl
WSO2 has made Agent Manager generally available, an open-source control plane for governing AI agents across any framework, model, or deployment. Licensed under Apache 2.0 and offered self-hosted or as managed SaaS, it covers agent identity, MCP governance, guardrails, sandboxed runtime, observability, and evaluation.

WSO2 has announced the general availability of Agent Manager, an open control plane built to govern AI agents across any framework, model, or deployment. The launch comes as enterprises struggle to keep pace with the rapid growth of automated AI agents in their operations.
The platform is fully open source under the Apache 2.0 licence and can be deployed anywhere, self-hosted or as managed SaaS. WSO2 frames it around two questions: what agents can access, and how agents run.
On the access side, Agent Manager governs large language models, MCP servers, and agent-level guardrails. On the runtime side, it provides a sandboxed runtime, observability, and evaluation.
Agent Manager entered beta in June 2026. General availability adds per-agent, per-environment identity controls, MCP-level governance, and a sandboxed runtime, pulling the full agent lifecycle into a single system.
WSO2's key architectural bet is separating governance infrastructure from agent logic, so teams can switch models or frameworks without rebuilding compliance controls. Chief AI Officer Dr Rania Khalaf said teams want the freedom to use the best model or framework for the job at hand, and that "control has to respect that heterogeneity. When governance is separated from agent logic, it can scale across frameworks instead of being locked into one ecosystem."
That framing targets a market problem now taking shape. Once enterprises move from pilots to production, the number of agents, their permissions, and the chains of calls between them expand quickly, and agents scattered across teams become a governance blind spot. Consolidating identity, permissions, MCP access, and runtime isolation into one control plane is one of the main industry answers to agent sprawl.
The open-source approach is worth watching. Apache 2.0 plus self-hosting means companies can keep the governance layer inside their own infrastructure rather than entrusting agent permissions to an outside vendor, but it also means WSO2 needs a commercial path beyond the open core, which is where managed SaaS comes in.
Two things to watch next: whether Agent Manager stays compatible with the main agent frameworks and the MCP ecosystem, and whether enterprises are willing to converge cross-department agent governance onto a single control plane.
Why it matters
As agent counts explode, consolidating identity, permissions, MCP access, and runtime isolation into one open control plane may become a prerequisite for production-scale agent deployments; framework compatibility and a viable commercial model decide whether it sticks.
Nearby Updates
All09/16, 14:02
vivo previews BlueCode phone coding agent, explores 30B MoE model for on-device use
vivo has previewed BlueCode, a coding agent built for phones, and says it is exploring a roughly 30B-parameter mixture-of-experts model for on-device use, according to TechNode. Both items are framed as previews and exploration rather than shipping products, with no launch date or supported handset list disclosed.
09/16, 13:00
University of Manchester Uses NVIDIA Earth-2 to Forecast UK Air Pollution
The University of Manchester is using NVIDIA's Earth-2 platform to forecast air pollution across the United Kingdom. With air pollution linked to an estimated 30,000 deaths in the U.K. last year and chemistry-based models costly to run, data-driven forecasting is drawing fresh attention.
09/16, 13:00
Tencent Cloud Open-Sources Its In-House AI Assistant Octop
Tencent Cloud has officially open-sourced Octop, its in-house AI assistant. The move turns the assistant from a vendor-maintained product into a project outside developers can inspect, deploy, and build on.
09/16, 15:00
Cohesity launches Agent Resilience to protect and recover AI agent infrastructure
Cohesity has introduced Agent Resilience, a capability aimed at protecting and recovering the infrastructure that AI agents run on. The announcement focuses on the agent runtime rather than the models themselves, and does not disclose pricing, availability dates or supported platforms.