Nintex adds on-premises AI to K2 for regulated firms

Nintex has released version 5.9.1 of its on-premises K2 business orchestration platform, adding a locally hosted AI engine for organisations in regulated environments.

The release also updates identity management and accessibility in a product used to manage business processes in complex, case-driven settings. The AI functions run inside a customer's own environment rather than through external APIs or cloud services during normal operation.

The update addresses a longstanding constraint for organisations in sectors such as government, financial services and healthcare, where sensitive workflow data often cannot be sent to external providers. That has limited the use of many AI tools built around cloud software and third-party models.

In the latest version, users can add AI functions to forms and workflows through built-in actions such as sentiment analysis and severity scoring. These features can support decisions including routing, prioritisation and issue detection while keeping operations within existing governance controls.

In practice, customers can introduce AI into established workflows without moving core process data outside their own infrastructure. Nintex is positioning the release as an option for organisations that want to test and apply AI in tightly governed settings rather than rebuild systems around external services.

Niranjan Vijayaragavan, chief product and technology officer at Nintex, outlined the company's rationale for the release.

"Organisations have been under pressure to adopt AI, but for many, especially in regulated environments, the barrier hasn't been interest, it's been how to apply it responsibly," said Niranjan Vijayaragavan, chief product and technology officer at Nintex. "K2 (5.9.1) is our first step in bringing AI directly into the platform in a way that fits how our customers operate today. By keeping AI within the boundaries they already trust, we're giving teams a practical entry point to start using AI in workflows where it adds value without disrupting the control and governance their processes depend on."

Identity changes

Alongside the AI addition, K2 5.9.1 updates identity federation. The release introduces automated onboarding for OIDC-compatible identity providers, a guided setup process and built-in synchronisation intended to reduce manual administration across systems.

Identity management has become a practical challenge for large organisations running multiple applications and access policies. By automating more of that work, the update aims to reduce the burden on IT teams managing user records, permissions and sign-ins across business systems.

The platform update also includes accessibility and usability changes, with WCAG runtime improvements for forms covering contrast, zoom behaviour and focus states.

A new high-contrast style profile has been added, and the Workflow Designer now gives users greater control, including optional auto-save behaviour. These changes are intended to help both end users and developers working with process forms and workflow design tools.

Regulated sectors

The significance of the release lies in how it addresses a divide in the AI market. Much of the software market has centred on cloud-native AI services, but organisations with strict data sovereignty and compliance requirements have had fewer options when they need to keep data processing fully on premises.

For those customers, the question is not simply whether to adopt AI, but whether the deployment model meets internal governance standards. In that context, a locally hosted AI engine offers a way to use AI for limited, specific tasks within existing controls rather than shifting wholesale to external platforms.

K2 has long been used as an orchestration layer for structured and case-driven work. With version 5.9.1, Nintex extends that model by combining deterministic workflows with AI-based decisioning inside a governed environment.

The approach is intended to let organisations introduce AI incrementally, maintain visibility over automated decisions and support compliance where data sovereignty requirements are strict. Existing customers can obtain the new version through standard upgrade processes.

The update reflects a broader trend in enterprise software, as suppliers increasingly tailor AI products to operational constraints rather than assume customers can rely on public cloud services. In regulated industries, that distinction may determine whether AI moves from pilot projects into day-to-day workflow systems.