Key Takeaways:
- IBM is expanding watsonx Orchestrate to support faster development and deployment of AI agents across departments.
- Business users can now build agents in minutes through no-code tools, while developers gain deeper access via pro-code kits.
- Multi-agent orchestration allows AI agents to work together, across frameworks and platforms, to automate complex workflows.
- New governance features provide observability, lifecycle control, and policy enforcement at scale.
- IBM reports $3.5 billion in internal productivity gains and a 176% return on investment from agentic AI rollouts.
IBM is deepening its investment in agentic AI systems with major upgrades to watsonx Orchestrate, introducing tools to help businesses scale autonomous digital workers across departments while maintaining governance and control. The announcement reflects a strategic shift: from experimental AI prototypes to production-grade, multi-agent ecosystems designed for real business outcomes. A post from Ritika Gunnar, General Manager, Data and AI, IBM gave great insight on what’s happening at Big Blue.
At the center of IBM’s latest update is a commitment to accessibility and speed. The new Agent Builder tool allows non-technical users to build agents in less than five minutes through a no-code interface. These agents can be created by combining preconfigured skills—such as answering emails, checking inventory, or initiating approvals—with enterprise data connectors. IBM has seeded the platform with over 150 pre-built templates covering common sales, HR, IT, and procurement tasks.
For more complex or custom implementations, IBM has introduced a pro-code Agent Development Kit. This gives developers the ability to build more tailored, model-rich agents using Python and open frameworks. It also supports integrations with LangChain, CrewAI, LangGraph, and other orchestration libraries, enabling advanced agent behavior such as multi-step reasoning, dynamic tool selection, and context-aware planning.
What distinguishes IBM’s approach in this wave of agentic AI is its focus on multi-agent orchestration. Rather than isolate agents to single tasks or business units, watsonx Orchestrate is designed to coordinate multiple agents—built internally or externally—across diverse workflows. A procurement agent may detect a delayed shipment, then trigger an inventory agent to update forecasts and a finance agent to delay payment. These agents can operate asynchronously or in parallel, exchanging context and updating shared data layers in real time.
To enable this kind of coordination, IBM is adopting the Model Context Protocol, which allows agents to share structured information and persist state across tools and interactions. This protocol ensures continuity even as workflows move between AI models, human users, and third-party systems.
But with scale comes risk. IBM’s latest update introduces an expanded set of governance and observability tools designed to keep agent behavior aligned with business rules, security protocols, and compliance mandates.
These include:
- Agent discovery and inventory, allowing IT teams to maintain visibility into which agents are active, their source, and what systems they interact with.
- Usage dashboards that track task volume, latency, and outcomes, helping operations teams optimize performance.
- Policy enforcement mechanisms to ensure agents adhere to internal data rules, approval structures, and usage limits.
- Lifecycle controls for versioning, deprecation, and rollback—essential in environments with dozens or hundreds of autonomous actors.
- Security context alignment, where agents inherit role-based access from connected applications, limiting exposure and enforcing Zero Trust principles.
One example of this security-first approach is AskIAM, a tool that enables AI agents to request access rights or identity verification across platforms like Slack or Microsoft Teams. Rather than hardcoding access logic, agents dynamically request permissions through IAM workflows, mirroring how humans request escalations or role changes.
IBM’s enterprise footprint makes it uniquely positioned to integrate agentic AI into existing enterprise software environments. The platform currently supports over 80 applications including Salesforce, SAP, ServiceNow, Workday, Microsoft 365, and Oracle—allowing agents to operate across a full range of enterprise functions without custom middleware.
This level of integration and interoperability is essential as AI agents move from back-office tasks to customer-facing roles and strategic processes. It also reduces the friction typically associated with AI implementation—no new data warehouses, no rip-and-replace architecture, and no vendor lock-in.
To validate its strategy, IBM points to internal deployments. Acting as its own testbed, the company reports over $3.5 billion in productivity gains across more than 70 business functions, including IT support, finance operations, human resources, procurement, and sales. In HR alone, agents now resolve 94% of simple employee requests without human intervention. In IT, ticket deflection has reportedly reduced workloads by up to 70%.
In a commissioned study, IBM claims that hybrid AI deployments driven by watsonx Orchestrate have achieved a 176% return on investment over three years, measured through time saved, task automation, and process acceleration.
For enterprises, this shift from AI experimentation to structured, governed deployment is a critical inflection point. Leaders increasingly want automation that delivers measurable value—but they also want control, trust, and scalability. Agentic AI, in IBM’s framing, is not a single product but an ecosystem that needs to be deployed carefully and intentionally, with oversight mechanisms built in from the beginning.
As organizations begin deploying more agents, the challenges of sprawl, security, and transparency will increase. IBM is betting that it can win trust by providing the orchestration layer and governance foundation enterprises need to manage that complexity.
The strategy also reflects broader industry trends. With major players like Microsoft, Google, and Amazon building agent frameworks atop their cloud platforms, the battle is no longer just about model performance. It’s about enterprise readiness—who can deliver a coherent, safe, and flexible system for companies looking to embed AI agents deep into their operations.
In IBM’s case, that means staying aligned with open standards, giving customers flexibility in how and where they build agents, and reinforcing that AI adoption can scale without sacrificing security or business control.
The new tools in watsonx Orchestrate represent a meaningful advance for companies ready to go beyond pilots. By enabling teams to deploy, connect, and govern AI agents with both speed and discipline, IBM is making a strong play for the operational layer of the AI enterprise stack.
Learn how AI Agents can supercharge your company’s profits and productivity at TMC’s AI Agent Event in Sept 29-30, 2025 in DC.
Rich Tehrani serves as CEO of TMC and chairman of ITEXPO #TECHSUPERSHOW Feb 10-12, 2026 and is CEO of RT Advisors and is a Registered Representative (investment banker) with and offering securities through Four Points Capital Partners LLC (Four Points) (Member FINRA/SIPC). He handles capital/debt raises as well as M&A. RT Advisors is not owned by Four Points.
The above is not an endorsement or recommendation to buy/sell any security or sector mentioned. No companies mentioned above are current or past clients of RT Advisors.
The views and opinions expressed above are those of the participants. While believed to be reliable, the information has not been independently verified for accuracy. Any broad, general statements made herein are provided for context only and should not be construed as exhaustive or universally applicable.
Portions of this article may have been developed with the assistance of artificial intelligence, which may have contributed to ideation, content generation, factual review, or editing.
