ChatGPT Gets ServiceNow Integration to Act on Work Tasks

OpenAI has expanded its collaboration with ServiceNow, the US provider of enterprise workflow and IT service management platforms. Users will be able to reach work information and carry out certain actions in ServiceNow directly from a ChatGPT conversation.

The change moves the assistant from giving answers to taking concrete steps inside company systems, within the permissions it is granted.

What the integration does

According to OpenAI's announcement, an IT employee will be able to search for similar incidents, assess the risk of a planned change and then update a ServiceNow record without leaving ChatGPT. Customers choose which capabilities the assistant may use.

ServiceNow applies the organization's existing rules and employee permissions. A user therefore cannot trigger an action they would not be allowed to perform in the original system.

What the companies plan next

The two companies plan to connect Codex to application development tools. They also intend to use OpenAI models to scan for security risks, rank vulnerabilities by priority and help coordinate fixes.

A voice interface is also in development. An employee could describe a use case by speaking, and the system would gather the required details and fill in gaps by asking follow-up questions.

Why it matters

Employees would no longer need to open several systems, search each one and copy data between them by hand. But the ability to execute actions makes permission control more important. The setup relies on ServiceNow's own rules rather than giving ChatGPT open-ended access, a model that could become a template for how AI agents are used inside companies.