October 27, 2026. At TechXchange 2026, IBM (NYSE: IBM) is introducing new capabilities that help developer, IT, and security teams deploy AI agents in enterprise systems in a controlled and traceable manner. IBM now combines agent detection, verified identities, access controls, and performance evaluation so that organizations can understand who owns an agent, which resources it is authorized to access, and what actions it performs across existing applications and infrastructure.
Categories
Tags
Contact info
Silicon Saxony
Marketing, Kommunikation und Öffentlichkeitsarbeit
Manfred-von-Ardenne-Ring 20 F
According to Gartner, the average Fortune 500 company will deploy more than 150,000 AI agents by 2028. Nevertheless, only 13 percent of companies believe they have the right governance in place for AI agents [1]. Persistent access credentials, unclear responsibilities, and siloed control mechanisms make it difficult to manage agents as they interact with other agents, use tools, and operate across teams, platforms, and cloud environments.
“The winners in enterprise AI are transforming the way their companies operate, rather than simply adding more AI,” says Dinesh Nirmal, Senior Vice President, IBM Software. “For agents to take on productive tasks across a company’s business-critical systems, an AI operating model is needed. A model that connects agents to data, applications, and infrastructure and operates them with the same discipline as the company’s most critical systems. When trust and control are built in from the start, agents become more than just a technology—they become a new way of doing business.”
Today’s announcements expand the IBM portfolio with an AI operating model based on four key pillars: agents, data, automation, and hybrid infrastructure. The starting point is trust and control for every agent—through a verified identity, exactly the access rights needed, and a central platform for detecting, securing, and evaluating agents, regardless of where they were developed.
HashiCorp Vault manages secrets and credentials that applications and services need to access enterprise systems. The solution offers centralized management of secrets, automatic rotation of supported credentials, and dynamically generated credentials that can be set to expire or revoked. These features reduce the time during which compromised credentials remain usable and restrict access to the permissions defined for a role.
HashiCorp Vault already protects non-human identities at companies and organizations such as ABN AMRO, Adobe, athenahealth, BBVA, BlackRock, Canva, Duke Energy, Expedia, the London Stock Exchange, Starbucks, Swiss Federal Railways, and Target—in environments that process billions of dollars in transactions [2]. In the future, HashiCorp will extend these capabilities to AI agents as well, assigning them secure, short-lived credentials that limit their permissions to what is strictly necessary.
AI agents represent a growing category of non-human identities, alongside applications and services. As they increasingly take on tasks within enterprise systems, organizations need a clear understanding of who owns an agent and what permissions have been granted to it. HashiCorp therefore combines managed agent identities with credential controls to help organizations delegate tasks with clearly defined responsibilities and permissions.
When used with HashiCorp Vault, Agent Identity serves as the authoritative source of information for each agent. This includes, for example, the owner, lifecycle status, and the permissions granted. The solution is designed to extend existing identity and security investments and pass relevant identity and authorization contexts to Vault’s credential controls. This enforces least-privilege access and revokes permissions in real time if an agent performs unauthorized actions.
Identifying an agent is only the first step. Organizations must also be able to control which credentials an agent uses for databases, applications, and cloud services—and how long those credentials remain valid.
When agents use tools and interact with corporate systems, their credentials become a critical security boundary. If an agent has long-term or overprivileged credentials, systems and data can be compromised if the agent is compromised or acts outside their intended scope of duties. Organizations must therefore define which tasks an agent is permitted to perform, restrict their access accordingly, and be able to revoke permissions as soon as circumstances change.
These controls go beyond initial authentication. If an agent attempts to retrieve information or perform actions that fall outside their approved permissions, access can be denied in real time and the session terminated. Vault’s short-lived credentials and least-privilege controls further limit the scope and duration of access rights.
Customers share their experiences with Vault at TechXchange.
“AI offers extraordinary opportunities to accelerate biomedical research and scientific discoveries. However, to realize this potential, security, trust, and responsible governance are essential,” says Shailesh M. Shenoy, Assistant Dean for Information Technology at the Albert Einstein College of Medicine. “In my experience, technologies such as IBM Verify and HashiCorp Vault can play a key role in providing organizations with the identity, access, and security controls they need to drive innovation with confidence.”
An agent’s identity tells organizations what an agent is. However, they also need to know which agents they have in total, regardless of the platform on which they were created. IBM watsonx Orchestrate operates cross-platform throughout the entire AI ecosystem, not just within a single vendor’s platform. This gives organizations a central location to discover, secure, and evaluate their agents. Features include:
In addition, Orchestrate works with Vault’s credential controls and Agent Identity to determine which tasks an agent is allowed to perform, what permissions it has, and how the principle of least privilege is enforced. Together, these controls help organizations authorize and audit agent actions, as well as quickly revoke trust and access credentials when risk landscapes change.
Once agents are authorized to access enterprise APIs and tools, they need to understand which resources are already available, how systems are interconnected, which components can be safely reused, and what impact their actions might have. The IBM Integration Intelligence Hub was developed to provide this knowledge through its Unified Context Engine. This enables agents, for example, to resolve stuck messages or provision APIs by leveraging existing resources. Capabilities, approval policies, and audit controls ensure that these actions remain controlled and traceable. Companies do not need to rebuild their IT landscape for AI; instead, they can deploy agents on the existing systems on which their business already depends.
Agent Identity and HashiCorp Vault integrate with the secure runtime environment NVIDIA OpenShell, the foundation of the NVIDIA OpenShell Architecture. Together, they provide a protected execution environment for agents with verified identities and short-lived credentials for each agent. This allows every action to be tracked, restricted, and audited.
“Companies need enforceable control over how agents operate and what they can access,” says John Fanelli, Vice President of Enterprise Software at NVIDIA. “Our collaboration with IBM combines the policy-driven runtime environment of NVIDIA OpenShell with Agent Identity and HashiCorp Vault to enable enterprises to scale AI agents—with secure execution, verified identities, and precisely limited access to business systems.”
When new vulnerabilities are discovered, enterprises must quickly identify which systems are affected, what changes are required, and how fixes can be implemented without disrupting critical business processes. In complex IT landscapes, this often involves multiple teams, tools, and processes.
IBM Concert integrates observability, optimization, incident management, and security across applications, infrastructure, and networks. Using AI-powered insights and evidence of the actual exploitability of vulnerabilities, Concert helps teams identify high-priority risks, assess operational impact, and coordinate countermeasures. At the same time, these capabilities help organizations improve performance, control costs, and strengthen the resilience of hybrid environments.
HashiCorp Terraform provides infrastructure context and controlled workflows for implementing changes. New features that are now generally available or available as a public preview—including Terraform Policy, the Terraform Developer Portal, and the Terraform Remote MCP Server—enable developers and AI agents to access approved configurations and implement infrastructure changes within established control mechanisms. Terraform, powered by Infragraph, also helps teams better understand infrastructure dependencies and relationships and make more informed decisions.
Together, these features enable the journey from identifying a vulnerability to implementing and verifying the fix, while providing the transparency and governance required for enterprise operations.
IBM Z Atlas maps a customer’s entire z/OS stack across infrastructure, middleware, and applications, and uses this visibility to simplify manual tasks such as patch management, regulatory compliance, and knowledge transfer between teams. This includes both system programmers and application developers. Atlas addresses the most common cause of unplanned z/OS outages: changes whose impact and security have not been sufficiently validated. Atlas analyzes how changes affect the z/OS stack and tests and validates them before implementation. This approach, currently available as a Technology Preview, accelerates customers’ modernization initiatives by putting control in the hands of experts while providing operations and development staff with the same transparent view of the Z system. This helps reduce risks.
HashiCorp Vault 2.2, Agent Identity, the Evaluate features of watsonx Orchestrate, IBM Concert, and HashiCorp Terraform Policy are now generally available. Secure Hub, the Terraform Developer Portal, the Terraform Remote MCP Server, and IBM Z Atlas are currently in public preview. The IBM Integration Intelligence Hub, as well as the Discover and Contain features of watsonx Orchestrate, are currently available as a technology preview.
Note: Statements regarding IBM’s future direction and plans are subject to change or withdrawal at any time without notice. They are intended solely to describe goals and intentions and do not constitute a binding commitment.
IBM TechXchange is IBM’s annual conference for developers and technology experts, bringing together nearly 6,500 attendees from 85 countries. This year, the event will be expanded to include HashiConf, thereby uniting the HashiCorp community with experts from the IBM, Red Hat, and Confluent ecosystems. Attendees can choose from more than 1,000 sessions and offerings, including 180 hands-on labs, 260 customer-led sessions, and more than 50 certification exams. Speakers include American Airlines, Bradesco, Citi, CVS Health, Deutsche Bank, DoorDash, Hertz, and US Bancorp, among others. They will discuss how they are leveraging AI productively and securely for enterprise use.
Live streams and recordings will be available here: ibm.biz/txc26-livestream.
[1]: Gartner Identifies Six Steps to Manage AI Agent Sprawl
[2]: HashiCorp Vault client stories
IBM is a leading provider of global hybrid cloud and AI solutions, as well as consulting services. We help clients in more than 175 countries leverage insights from their data, optimize business processes, reduce costs, and gain a competitive edge. Thousands of governments and enterprises in critical infrastructure sectors such as financial services, telecommunications, and healthcare rely on IBM’s hybrid cloud platform and Red Hat OpenShift to drive their digital transformation quickly, efficiently, and securely. IBM’s groundbreaking innovations in AI, quantum computing, industry-specific cloud solutions, and consulting provide our clients with open and flexible options. All of this is based on IBM’s longstanding commitment to trust, transparency, accountability, inclusion, and service.
– – – – –
Photo: IBM
You need to load content from reCAPTCHA to submit the form. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from Turnstile. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from Facebook. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from Google Maps. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from OpenStreetMap. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from Instagram. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from X. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information
