Skip to main content

WSO2 Launches Open-Source Agent Manager to Centralize AI Governance

18 SEPTEMBER 2026·2 MIN READ·3 SOURCES·Trusted source

WSO2 has announced the general availability of its open-source Agent Manager, a platform built to secure, monitor, and govern AI agents across diverse models and frameworks.

WSO2 Launches Open-Source Agent Manager to Centralize AI Governance

Key takeaways · 3

  • 01

    WSO2 Agent Manager separates agent governance infrastructure from underlying agent logic.

  • 02

    The platform includes over 40 built-in controls for policies like rate limiting and PII masking.

  • 03

    A Kubernetes-native sandboxed runtime provides isolated execution environments for enterprise agents.

Centralizing Agent Governance

WSO2 has announced the general availability of Agent Manager, an open-source platform designed to centralize governance and operational oversight for AI agents across different models and frameworks. [1] The platform separates agent governance infrastructure from agent logic, allowing organizations to maintain common controls when swapping out underlying models or frameworks. [1][3] Enterprise adoption of agents continues to grow faster than the infrastructure designed to govern them, with Gartner predicting that the average Fortune 500 company will utilize over 150,000 agents by 2028. [3] The WSO2 platform is built to operate across cloud, on-premises, and hybrid environments, avoiding lock-in to a single AI provider. [1]

Security and Identity Controls

Agent Manager initially launched in beta during June 2026, and the general availability version introduces new capabilities for agent identity and Model Context Protocol (MCP) governance. [1][3] The platform features a Kubernetes-native sandboxed runtime intended to provide isolated, controlled execution environments for agents. [1][2] To manage authority, Agent Manager includes a central management layer that handles verifiable agent identity, token exchange, access revocation, and role-based access. [1] It also provides more than 40 built-in controls that can be applied across agent, MCP, and LLM layers, including areas such as rate limiting and personally identifiable information masking. [1]

What it means

By separating governance from logic, WSO2 is addressing the friction of fragmented agent management across tools like LangChain, CrewAI, and Amazon Bedrock. Dr. Rania Khalaf notes that this decoupling allows teams to scale control mechanisms across frameworks without being restricted to a single ecosystem. As enterprises move past experimental agent development and face massive scale projections, centralized identity and execution sandboxes will become baseline requirements rather than optional add-ons. What the sources don't address: How the open-source Agent Manager integrates with or replaces existing API gateways and enterprise identity systems already in place.

Enterprises are rapidly scaling their use of autonomous AI agents, creating severe security and management blind spots. A unified control plane allows teams to enforce identity and policy guardrails across fragmented AI infrastructure without provider lock-in.

Why it matters
Daily session

Put this to work — one session a day, built for your industry.

Create a free account for a daily session — eight questions and one real-work challenge, on the news that affects your role.

Start free

How this developed

  1. 18 September 2026

    WSO2 Launches Open-Source Agent Manager to Centralize AI Governance

  2. 18 September 2026

    Event created from source cluster.

Sources

AI fluency, one session a day, built for your work.