OpenAI and ServiceNow have expanded their partnership so enterprise teams can carry out governed company workflows from inside ChatGPT and Codex, the companies announced on October 28, 2026.
The expansion introduces a new ServiceNow plugin for ChatGPT, a Codex integration for building applications, and joint work on cybersecurity and voice-driven risk assessment — a step both companies frame as moving enterprise AI from reasoning to governed action.
“AI becomes more useful when people can put it to work in the systems their businesses depend on,” said Greg Brockman, OpenAI’s president and co-founder. “By bringing our models together with ServiceNow’s workflows and business context, we’re helping teams build software, address security risks and put AI to work across their organizations.”
What the expansion actually adds
At the centre is a two-way connection. OpenAI models and agents gain access to ServiceNow workflows through ServiceNow’s Action Fabric, while ServiceNow data and workflow actions appear inside ChatGPT and Codex.
In practice, an IT employee could investigate a problem, find similar past incidents, assess the risk of a proposed change and update a ServiceNow record without leaving the ChatGPT conversation. Customers connect ChatGPT to their own ServiceNow instance and choose which functions to expose.
Governance is the key constraint: Action Fabric enforces the customer’s existing business rules and access controls, so the actions available to a given employee mirror that organisation’s own permissions. The plugin surfaces business context, incident handling, risk assessments and workflow actions inside ChatGPT, and is available in OpenAI’s marketplace, with enhanced capabilities expected by the end of 2026.
Codex, security and voice
For developers, a Codex integration in ServiceNow’s AI Workflow Factory will let teams build applications that inherit ServiceNow context and governance, rather than bolting AI onto workflows afterwards.
On security, ServiceNow’s Security and Risk products will integrate OpenAI cybersecurity capabilities to help teams identify, triage, validate and mitigate enterprise attack paths. Those features are scheduled for early access in January 2027, with general availability projected for the second quarter of 2027.
ServiceNow’s Super Intelligence Control Tower will use OpenAI voice models for more seamless AI risk assessments, with voice capabilities slated to launch by the first quarter of 2027. OpenAI is also joining ServiceNow’s newly announced Control Tower Alliance as a founding member.
Built on the January agreement — and adopted internally
The expansion builds on the multi-year strategic agreement the companies announced on January 20, 2026, when ServiceNow designated OpenAI models as a preferred intelligence layer for its customers, including GPT-5.2 on the ServiceNow AI Platform.
Both companies are also adopting the tools themselves: OpenAI plans to make ServiceNow available across its workforce, while ServiceNow plans to make ChatGPT Work and Codex available to all of its employees, expanding from engineering-only access.
No contract value, revenue contribution or overall implementation timeline was disclosed for the expanded partnership, and this article does not invent one. Availability also varies by feature: the ChatGPT plugin is live, while several security and voice features are dated for 2027.
Why it matters
The deal targets a long-standing gap in enterprise AI: frontier models can reason over text but usually cannot take approved action inside the systems where work lives, while enterprise software can route approvals but has often lagged on AI capability. By connecting the two under existing permissions, OpenAI and ServiceNow are betting that governed action — not just chat answers — is what will win corporate budgets.
Related Emadan coverage: Oracle Brings ChatGPT and Codex to Its Apps, OpenAI to Sell Its Top Models on Google Cloud, and OpenAI’s GPT-6.1 Terra Arrives on Amazon Bedrock as a Cheaper Computer-Use.
Frequently asked questions
What is the new ServiceNow plugin for ChatGPT?
It brings ServiceNow business context and workflow actions into ChatGPT conversations, so users can investigate issues and complete approved tasks without switching applications.
Can ChatGPT do anything in ServiceNow?
No. Actions are limited by the customer’s own ServiceNow permissions and business rules through Action Fabric, and customers choose which functions to expose.
When are the security and voice features available?
Security features enter early access in January 2027 (general availability projected Q2 2027); voice-driven risk assessment is slated for Q1 2027. The ChatGPT plugin itself is already available.
Is this a new partnership?
No — it expands a multi-year agreement from January 20, 2026. The October 28 announcement deepens it into a two-way product integration.
Sources
- BusinessWire — ServiceNow and OpenAI accelerate enterprise AI
- Unite.AI — OpenAI and ServiceNow Expand Partnership With ChatGPT Workflow Plugin
- Crypto Briefing — ServiceNow and OpenAI expand partnership with ChatGPT plugin and security features
- TokenPost — OpenAI and ServiceNow Expand Partnership Across Enterprise Workflows
