IBM Unveils Content Cortex

Published June 3, 2026 on novadoc.nl

IBM has introduced Content Cortex, a next-generation, AI-native content services platform that the company positions as the future of enterprise content management. Announced at IBM Think 2026 and currently available in preview via a waitlist, Content Cortex aims to solve a problem that has stalled many corporate AI initiatives: vast amounts of unstructured content that is locked away in silos, poorly governed, and out of reach for AI agents.

The evolution of FileNet and Content Manager

For organizations already invested in IBM's content stack, the most important news is one of continuity. IBM describes Content Cortex as the natural evolution of FileNet Content Manager, Content Manager OnDemand, and Content Manager Enterprise Edition — unifying the company's content services portfolio into a single, AI-ready platform. Rather than a rip-and-replace, IBM positions it as building on three decades of proven, enterprise-grade storage, governance, and scalability, now rebuilt for AI from the ground up.

According to IBM, a single system can process billions of documents, petabytes of data, and millions of API calls per day, each time with complete audit trails.

Three core ideas: centralize, govern, activate

Content Cortex is built around three capabilities. The first, Centralize , brings content from any system together into a single managed repository and enriches metadata at scale, so that documents are findable and consumable by both people and AI agents, while honouring document and user security. A single source of truth, no scattered versions.

The second, govern, integrates access control, retention policies (archiving), and audit trails by default into every document and its semantic representation, throughout the entire content lifecycle. IBM emphasizes that governance is part of the platform (governance by design) rather than something bolted on afterwards, with every action sourced, timestamped, and defensible.

The third capability, activate , is where the "agentic" promise comes in. Content Cortex lets AI agents execute content operations — classify, extract, redact, apply legal holds, and search — through native Model Context Protocol (MCP) integration. Importantly, IBM has built this to work with any AI runtime of the customer's choice, including watsonx Orchestrate, Claude, Copilot, and ChatGPT.

Features and an "Essentials" base

The platform's Essentials edition is presented as the governance foundation for an agentic enterprise. Headline features include intelligent archiving with up to 30:1 compression for long-term content and instant retrieval, a remote MCP server, a Navigator Core Agent that gives staff natural-language access to content directly inside Content Navigator, enterprise security and governance, Business Automation Insights dashboards, and automated retention and disposition policies.

IBM emphasises a "zero data leakage by architecture" approach: security is enforced at every layer, and when a document is deleted, all of its derived embeddings are deleted too.

Built for document-intensive organizations

Content Cortex is aimed squarely at document-intensive (regulated) sectors: banking and financial services, insurance, healthcare, government, and manufacturing. IBM points to customer case studies to make its case. BlueShore Financial reportedly cut a loan backlog from six weeks to overnight and grew lending volume 250% over three years without additional hires. Central Nacional Unimed achieved 90% lower administrative costs. Tatweer Building reduced the time to fulfil information requests from up to three months to under an hour.

Availability

Content Cortex is currently available as a preview. Organizations can sign up for the waitlist via the IBM product page at ibm.com/products/content-cortex. The platform is hybrid, as it supports on-premises , private cloud , and public cloud , and a SaaS option is planned for later in 2026.

Novadoc is happy to help demonstrate or implement the product. Novadoc also specializes in migrating from earlier versions of FileNet, CM, or CMOD to Content Cortex.