Skip to content
‹ Back to IndustriesSoftware & Technology

OpenAI and ServiceNow expand partnership to connect workflows with ChatGPT

According to OpenAI, ServiceNow's Action Fabric applies each customer's existing business rules and permissions to what people can do in ChatGPT.

This story was produced through MarketScale. See how Software & Technology teams put it to work with Executive Thought Leadership.

By MarketScale Newsroom · ServicenowOpenaiChatgptCodex
Share
OpenAI and ServiceNow expand partnership to connect workflows with ChatGPT

Key takeaways

01

ChatGPT inherits a ServiceNow instance's existing business rules and permissions through Action Fabric, so years of accumulated role settings now govern a new conversational interface.

02

The ServiceNow MCP plugin is available in the OpenAI marketplace now; joint OpenAI-ServiceNow security tooling enters early access in January 2027 and general availability in the second quarter of 2027, which puts it squarely in 2027 security planning.

03

With OpenAI now partnered with both IBM and ServiceNow, model access is a weak basis for comparing platforms; the permission enforcement and record context a model plugs into is the sharper thing to evaluate.

Free workspace

Turn your Software & Technology expertise into content.

Record interviews, organize footage, and write with AI on a free trial of the MarketScale platform for qualifying companies. No demo required, no credit card.

Try it Free

An on-call engineer opens ChatGPT, pulls up the monitoring alerts tied to an incident, checks what changed recently and which services are affected, then assigns the right team and starts on a fix. That is one example of the interactive experiences ServiceNow is previewing inside ChatGPT as part of the expanded partnership it and OpenAI announced on October 28, 2026. The partnership connects ServiceNow workflows with ChatGPT and Codex and puts more OpenAI models and agents inside ServiceNow products.

Part of it is already live. According to OpenAI's announcement, customers can get the ServiceNow MCP plugin from the OpenAI marketplace today, and enhanced capabilities are expected to follow by the end of 2026. The integration is built so that an IT staffer can look up similar incidents, gauge how risky a change is and update a ServiceNow record, all without leaving the chat.

For whoever owns the ServiceNow instance, a second front door raises an obvious question: what can people do once they walk through it? The announcement's answer is short. Whatever the instance already allows.

Action Fabric decides what ChatGPT can touch

Customers connect ChatGPT to their own ServiceNow instance and pick which capabilities to make available. ServiceNow's Action Fabric then applies the organization's existing business rules and permissions, so an action taken from ChatGPT reflects the same configuration and access rights as one taken in ServiceNow itself. OpenAI president Greg Brockman put the logic plainly in the announcement: AI gets more useful when it can act inside the systems a business already depends on.

That design moves the governance work to before the chat starts. The new front door into ServiceNow is ChatGPT, and the locks on it are the business rules and permissions an IT team configured long before anyone typed a prompt. For organizations with clean role definitions, that's convenient, since there's no second permission model to build and maintain. For instances where access rights have piled up through years of exceptions, the same inheritance would carry those exceptions into an interface that makes acting on a record faster and easier.

Before switching the plugin on broadly, check which ServiceNow roles carry write access to incident and change records. Action Fabric applies those same permissions to anything requested through ChatGPT, so the capability list chosen at connection time and the role model underneath it are the real controls.

The model reasons, the platform acts

The groundwork dates to January 20, 2026, when the companies announced a multi-year agreement making OpenAI models a preferred intelligence capability for ServiceNow enterprise customers. ServiceNow's release at the time said the models would work alongside a customer's configuration management database (CMDB). It also said its AI Control Tower would offer one central view covering three things: where models are used across workflows, the way they work with enterprise systems and data, and whether AI-driven actions are carried out in a controlled, auditable way.

OpenAI's own January post described the same principle from the opposite direction. With GPT-5.2 built into the ServiceNow AI Platform, the model pairs with the workflow engine, where it can access enterprise data and respect governance and permissions while moving work through approvals and updates. That agreement covered enterprises running more than 80 billion workflows a year on ServiceNow. In January the model came into ServiceNow. In October, ServiceNow came into ChatGPT, and the permission rule made the trip both times.

Paul Fipps, ServiceNow's president of global customer operations, laid out the architecture behind that rule in an interview with Diginomica this year. ServiceNow uses large language models for decisioning and reasoning, the part he called probabilistic, and customers can bring whichever model they prefer; he named Gemini, ChatGPT and Claude. The workflow that carries out the action is deterministic. "No guessing," he said.

When the LLMs get better, that part of the process gets better, but the execution part is where we innovate., Paul Fipps, president of global customer operations, ServiceNow, to Diginomica

Read against that split, Action Fabric looks like the deterministic half of the ChatGPT integration: the chat can help someone reason through an incident, while the record update runs through rules the customer already set. It also suggests why ServiceNow can open a door for OpenAI without staking its platform on one model. Diginomica argued that treating the model as interchangeable insulates ServiceNow from model obsolescence in a way that vendors betting on a single provider are not. ServiceNow chairman and CEO Bill McDermott framed the October pairing in matching terms, as "turning world-class reasoning into trusted execution at global scale."

Dates that run into mid-2027

Most of the announcement is a schedule. OpenAI laid out the sequence this way:

  1. Now: ServiceNow MCP plugin available in the OpenAI marketplace
  2. By end of 2026: enhanced plugin capabilities
  3. January 2027: early access to OpenAI-powered security capabilities
  4. First quarter of 2027: voice in ServiceNow's Super Intelligence Control Tower
  5. Second quarter of 2027: general availability of the security capabilities
  6. 2027, no quarter given: Codex in AI Workflow Factory

What the security, developer and voice plans add

According to OpenAI, the companies will offer a joint security solution that pairs OpenAI's cybersecurity capabilities with the asset, identity and business context held in ServiceNow. The goal is to help teams investigate vulnerabilities, prioritize risks and coordinate remediation.

A security operations leader drawing up tooling plans for 2027 could use early access, which opens in January 2027, to test on the organization's own data whether the extra context improves triage. That window opens a few months before general availability, which is scheduled for the second quarter of 2027. Of all the pieces, identity has the longest visible track record. Fipps told Diginomica he prefers to deal with risk before anything else, move as quickly as that risk allows, measure again, and only after that go to production. The publication linked this approach to the way ServiceNow worked toward its AI Control Tower and its acquisition of Veza for identity security.

For developers, a Codex integration is planned in AI Workflow Factory that draws on ServiceNow's business context and platform capabilities. OpenAI models will also power Build Agent and Autonomous Engineer for application development and modernization. Customers already had related uses. ServiceNow's January release named tools for developers and admins that convert intent into automation, logic and workflows as one of the long-standing ways customers could access OpenAI models. The same release described OpenAI computer-use models enabling automation of legacy systems, including mainframes, which puts the modernization half of Autonomous Engineer's brief in context.

Code is also the area where Fipps's risk-first sequence pushes hardest. He described a meeting with a customer for whom writing code quickly with a large language model and then deploying it into a system of that kind is not something they would even consider. For development teams that share that caution, the part of the planned Codex integration to test first may be its dependence on ServiceNow's business context and platform capabilities.

In the Super Intelligence Control Tower, an OpenAI-powered voice experience will allow people to speak an AI use case out loud. The feature fills in assessment fields and asks follow-up questions to collect any information that is still missing. In January, ServiceNow said it was working with OpenAI toward real-time speech-to-speech AI agents. Its example had a user speaking in their preferred language while an agent, with no delay for translation, opened a case, set off an approval and orchestrated the steps that followed. The new feature looks more like a spoken intake form.

ServiceNow tests it on its own staff

ServiceNow plans to roll out ChatGPT Work and Codex to every employee, not only to its engineering organization. Its engineers can explore codebases, make changes and run tests with Codex, while other teams rely on ChatGPT Work for daily tasks. ServiceNow will also apply OpenAI capabilities to its internal security work, and OpenAI says it runs ServiceNow workflows in its own operations, supported by the Super Intelligence Control Tower.

Internal results are the evidence ServiceNow points buyers to. Fipps told Diginomica that the company no longer has a level one help desk because agentic AI now handles that work autonomously, and that many people who held those roles moved into cloud operations, level two support or data centers. Because ServiceNow is growing, he said, it wasn't a big layoff. Diginomica noted that enterprises growing less rapidly may find that argument harder to make for their own level one workers.

ServiceNow's internal benchmark, and the gap it targets

>90%
Share of ServiceNow employee IT requests now handled autonomously, per the company
99% faster
How quickly ServiceNow says its L1 Service Desk AI Specialist resolves assigned cases versus human agents
21.4%
Share of 35 CIOs and CTOs in Diginomica's November 2025 research reporting AI initiative success rates above 80%
40%
Negative sentiment in Diginomica's January 2026 pulse on 2025 AI projects, mostly about reaching production

Diginomica

Fipps is selling customers on speed. ServiceNow wants to halve the time to a first AI use case, from six months to three, by having its services team use AI tools to turn statements of work into project plans, design specifications and configuration. He said work that once took weeks or months now takes a week. That goes after the problem Diginomica's research keeps finding: projects that get stuck between proof of concept and production.

Diginomica's own assessment was that whether three months holds beyond early adopters with high technology maturity remains open. The vendor that builds a platform is also its most favorable test site, so comparable results from a heavily customized customer instance would say more about what the ChatGPT integration can do for everyone else.

OpenAI's partner list keeps growing

ServiceNow isn't OpenAI's only enterprise channel. In August, Lynn Greiner reported in Network World that IBM will embed OpenAI frontier models and products such as Codex and ChatGPT into IBM's AI consulting services platform, IBM Consulting Advantage. IBM will also join the OpenAI Partner Network, create a dedicated OpenAI practice that certifies thousands of consultants and engineers through the Partner Network, and expand an existing cybersecurity collaboration that combines OpenAI frontier AI capabilities with IBM Autonomous Security.

That deal sits inside a broader race to put forward-deployed engineers next to enterprise customers. OpenAI announced its OpenAI Deployment Company on May 11, a week after Anthropic launched its own effort, and Microsoft and AWS have joined in too. The IBM partnership also plans industry-specific products for financial services, government, telecommunications and retail.

Info-Tech Research Group distinguished analyst Mark Tauschek said deals like IBM's are "becoming table stakes" and "certainly not differentiating." He noted that enterprises will choose the consulting firms that work with their AI vendor of choice, gave Microsoft the initial advantage for quickly supplying thousands of skilled forward-deployed engineers, and called it "a horse race." Applied here, that could be the strongest objection to reading much into the ServiceNow news: if every major platform has frontier models, access to them separates no one.

Fipps's answer to a related challenge, OpenAI and Anthropic moving into enterprise workflow territory, was that working at the core of a business is "vastly different from vibe-coding a point solution." Diginomica pointed out that Salesforce, SAP and Workday each make some version of the control-layer argument too, and that about 80% of respondents in its CIO network survey were sustaining or increasing SaaS spend, with no definitive view on AI replacing the platform itself.

This argument shows which parts of the deal would be hardest for others to copy. For ServiceNow customers, the likely differentiators are Action Fabric's permission enforcement and the incident, change and asset records their instances already hold. That platform data could count for the most in security. Diginomica characterized ServiceNow's wider bet as a hope that, as AI becomes more agentic, the company that owns the layer where execution and governance happen will capture most of the long-term value.

The first checkpoint is the enhanced ServiceNow MCP plugin, which OpenAI expects by the end of 2026. For operators who saw the on-call preview, it could show how much of that workflow ships as product. The second is the January 2027 early access for the OpenAI-powered security capabilities, where security teams could learn whether ServiceNow's asset and identity context improves vulnerability work on their own systems.

Featured companies

Your experts belong here

Every story in MarketScale Software & Technology starts with a company putting its solutions engineers, product teams, and customer engineers on the record. Buyers are already reading this topic. The only question is whose experts they find.

Buyers ask AI engines who to consider, and published expert answers are what those engines cite.

Book DemoSee how it works15 minutes, straight to a calendar.

About the author

MarketScale Newsroom
MarketScale NewsroomEditorial Team, MarketScale

The MarketScale Newsroom reports on the companies, technologies, and trends shaping 16 B2B industries. It turns primary sources and expert commentary into clear, useful coverage for the people doing the work.

B2B Weekly

The week in Software & Technology, and sixteen other industries, every Monday.

Ten stories, one-line takes, five minutes. Free.

Software & Technology: are you visible to AI?

Before they reach out, Software & Technology buyers ask AI engines which vendors to trust. Explore how your experts, customers, and partners can become useful content for buyers and AI search.

Free Trial

You just read one Software & Technology expert. Your company is full of them.

This article was produced through MarketScale. The same platform turns your solutions engineers, product teams, and customer engineers into the articles, video, and social content Software & Technology buyers are searching for. Start a free trial and see it with your own people. For qualifying companies, no credit card, no demo required.

NPS +73 · 1,000+ creators · 38+ countries

What your free trial includes

Hands-on access to the MarketScale platform
Media requests to your crowd, remote recording, AI writing tools
No demo required. No credit card.
For qualifying companies. Company confirmation required.

More Software & Technology Insights

HFS survey: only a third of AI-active enterprises hit their AI business case

HFS survey: only a third of AI-active enterprises hit their AI business case

Only 33% of 502 senior AI decision makers at AI-active enterprises said their biggest AI program met or beat its business case in an HFS Research–Rackspace survey. Success drops to 28% when AI spans several functions. HFS says the pattern points to a fixable gap: unclear ownership of live AI performance, cost and governance as one system.

  • 01Business-case success falls from 37% to 28% as AI moves from one function to several, and reaches 56% where AI is embedded in standard operations. The groups differ, so this is a pattern rather than proof.
  • 02Survey buyers signaled willingness to pay a premium for a single accountable operator—73% said they would pay for a contracting model tied to clear operational responsibilities and measurable outcomes. HFS says the appetite is strongest among companies that have already felt the ownership gap.
  • 03Before adding another function to a working AI program, settle who answers for its performance, cost and governance after go-live, and how each is measured.

Nov 15, 2026

IBM and Microsoft's Imagine Enterprise starts with more than 6,000 pre-built AI agents and assets

IBM and Microsoft's Imagine Enterprise starts with more than 6,000 pre-built AI agents and assets

Imagine Enterprise, the agent platform from IBM and Microsoft, pairs IBM Consulting's more than 6,000 AI agents and assets with Microsoft's AI, data, cloud and security foundation. Forward-deployed engineers from both companies support qualifying engagements. For regulated firms, a useful test could be how quickly one named process reaches production.

  • 01Imagine Enterprise combines more than 6,000 AI agents and assets, Microsoft's identity, security and AI governance tools, and forward-deployed engineers from both companies for qualifying engagements, which suggests buying it looks closer to a delivery engagement than a software download.
  • 02Microsoft products including Entra, Purview and Defender provide the identity, AI governance and security foundation, which could matter most for regulated firms already standardized on Microsoft.
  • 03The sharper pilot question is wall-clock completion time and success rate on one named process as more agents run at once, a yardstick Arm is pushing for agentic workloads.

Nov 15, 2026

Google’s TPU 8tx and 8i Max: bandwidth vs. confidential inference

Google’s TPU 8tx and 8i Max: bandwidth vs. confidential inference

Google Cloud has added two systems to its eighth-generation TPU lineup. Network World reports that TPU 8tx is Google Cloud’s first TPU with confidential computing, developed with AMD, and Zeniteq says it uses TPU 8t silicon for secured inference. Google says TPU 8i Max packages two TPU 8i chips together, doubling bandwidth and SRAM while increasing HBM capacity by about a third.

  • 01Neither system uses a new accelerator design: TPU 8tx pairs TPU 8t silicon with fifth-generation AMD EPYC hosts and Google's Boardfly network, while TPU 8i Max packages two existing TPU 8i chips together.
  • 02TPU 8i Max doubles memory bandwidth and on-chip SRAM but raises HBM capacity by 33%, and bandwidth and capacity answer different questions: how fast context data moves versus how much of it the system can hold.
  • 03The 2x latency claim for TPU 8tx is measured against TPU 8t, while the 8i Max gains are measured against TPU 8i. Neither figure says which system finishes a given job more cheaply.

Nov 14, 2026

Explore More Software & Technology Insights

Read more expert perspectives from across Software & Technology.

Browse Software & Technology Hub

About the Expert

MarketScale Newsroom
MarketScale Newsroom

Editorial Team

MarketScale

The MarketScale Newsroom reports on the companies, technologies, and trends shaping 16 B2B industries. It turns primary sources and expert commentary into clear, useful coverage for the people doing the work.

For B2B teams

Your experts could be publishing here

Stories like this one run on content MarketScale captures from real practitioners. See how your team's expertise becomes coverage in Software & Technology and beyond.

Book a Demo

Or call us. No forms required. We pick up. 214-945-2512