WSO2 has announced the beta launch of WSO2 Agent Manager, an open control plane built to help enterprises manage artificial intelligence (AI) agents across environments. The move comes as companies shift from early AI experiments to full-scale deployments, where concerns around governance, security, and operational control are becoming harder to ignore.
The platform aims to give organizations a single system to identify, monitor, and manage AI agents, which often operate autonomously and across multiple tools, frameworks, and cloud environments.
Growing adoption meets rising risks
Enterprises are investing in AI agents to unlock productivity gains, but operational readiness has not kept pace. Many teams face a trade-off between speed and control. Moving fast often leads to weak oversight, while building governance frameworks from scratch can slow progress.
Industry data reflects this tension. Gartner estimates that more than 40% of agentic AI projects could be canceled by 2027 due to high costs, unclear business value, and poor risk management.
Rania Khalaf, Chief AI Officer at WSO2, highlighted the challenge in a statement: “AI agents introduce a fundamentally new challenge. Their autonomy and probabilistic behavior make them powerful but also difficult to control.”
She added that the new platform is designed to bring agents into the enterprise system as “identified, governed, accountable entities” that can operate securely at scale.
Tackling agent sprawl
As organizations deploy more agents, a new issue is emerging—“agent sprawl.” This refers to a growing number of loosely coordinated agents, often running across different systems with inconsistent policies and limited oversight.
WSO2 Agent Manager attempts to address this by creating a centralized system of record for all agents. This allows enterprises to maintain visibility and apply consistent governance, regardless of where the agents are running.
The platform supports agents across cloud, on-premises, and hybrid setups, while allowing teams to continue using different frameworks based on their needs.
Key features and capabilities
WSO2 Agent Manager includes a set of features focused on control, transparency, and interoperability:
- Federated agent management: A single control plane to manage agents across multiple environments
- Agent identity and access control: Each agent is assigned a verifiable identity, with secure access delegation and audit trails
- Centralized governance: Policy enforcement across agents, large language models, and tools
- Observability and tracing: Detailed insights into agent behavior and performance
- Secure runtime environment: Built on Kubernetes with zero-trust principles and real-time intervention controls
- Open architecture: Based on standards like OpenTelemetry and OpenAPI, supporting frameworks such as LangGraph, CrewAI, and Ballerina
The platform is designed to avoid vendor lock-in, offering flexibility through its open-source Apache 2.0 license.
Looking ahead
WSO2 Agent Manager is expected to reach general availability in June 2026. The company positions it as an extension of its broader role in enterprise infrastructure, now adapted for AI-driven systems.
As businesses continue to explore agent-based automation, tools that balance innovation with accountability are likely to gain importance. The success of such platforms will depend on how well they integrate into existing workflows while addressing growing concerns around compliance, cost, and control.