Every enterprise that has spent the last two years adopting AI has some version of the same story. A summarization tool inside the help desk. A copilot bolted onto the CRM. A chatbot that answers HR questions until it hits the edge of what it knows, then hands the person back to a form. Ask an employee to actually get something done through any of these tools, and the story changes fast. They still open five different systems, still chase an approval over email, and still explain the same request three times to three different assistants that forget each other exist the moment the browser tab closes.
ServiceNow built Otto to close that gap. Announced on May 5, 2026 at Knowledge 2026 in Las Vegas, ServiceNow Otto is the company's new unified AI experience, a single conversational front door that sits across the entire ServiceNow AI Platform and finishes work instead of just describing it. For any organization already running ServiceNow, or evaluating it as a platform, Otto changes what "AI inside ServiceNow" actually means.
Here is what Otto is, how it works, where it stands today, and what has to be true on your own instance before it delivers on what ServiceNow is promising.
What is ServiceNow Otto?
Source: servicenow.com/in/platform/otto.html
ServiceNow sums up Otto in one line: employees, partners, and customers ask, and ServiceNow Otto handles the rest. That sentence is the whole pitch, and it also marks the difference between Otto and most of the AI tools enterprises have deployed so far.
Otto does not live inside a single application or answer a single type of question. It sits across ServiceNow's platform, works out what a request actually means in plain language, decides which system and which workflow owns it, and carries it through to completion. That might mean filing a request, running it through an approval chain, or pulling an answer from three different data sources at once.
Nenshad Bardoliwalla, ServiceNow's group vice president of AI product management, describes Otto's job as turning "intent into enterprise work for every person and across every workflow." Bhavin Shah, SVP and GM of Employee Experience and AI at ServiceNow, puts the practical effect more simply: employees "no longer need to know where to go or who to ask."
The completion problem
Most enterprise software vendors, including some of the biggest names in productivity and CRM software, build AI inside their own applications. That AI can summarize a document or draft a reply within that one app. It cannot reach into the four or five other systems a real request touches, because it was never wired into the approval chains, permissions, and audit trails a business process actually runs on.
A large language model on its own does not solve this either. A model can reason about what needs to happen, but reasoning about a request and executing it inside a governed system of record are two different jobs. The result is AI that answers questions but cannot finish them. ServiceNow calls this a completion problem, and the term fits: employees keep toggling between applications, approvals keep getting chased down by hand, and AI spending keeps climbing while the productivity gains many enterprises expected are slow to show up.
Otto runs on the same platform that already owns the workflow, approval, and permission logic for the business, so it is not guessing how to finish a request. It hands the request to a system that already knows how to execute it.
How ServiceNow built Otto
Otto did not appear from nowhere. It is ServiceNow folding three previously separate capabilities into a single experience.
Now Assist is ServiceNow's native generative AI, built to summarize, draft, categorize, and assist inside existing workflows.
Moveworks is the conversational AI company ServiceNow acquired for $2.85 billion, in a deal that closed in December 2025. Moveworks brought a front-end AI assistant, enterprise search, and a reasoning engine already connected to more than 100 enterprise systems.
AI Experience is the multimodal, agentic interaction layer ServiceNow previewed at Knowledge 2025, designed for orchestrating requests across systems rather than inside just one.
“Moveworks understood what employees needed. ServiceNow could do the work. Together, we built ServiceNow Otto, an AI experience that completes work, across any system, department, or any workflow.”
— Bhavin Shah, SVP and GM of Employee Experience and AI, ServiceNow
In practice, that means Otto does not route a request to Now Assist or Moveworks depending on which team built which feature. It decides, based on the request itself, whether the right response is a quick generative answer, a multi-step agentic workflow, or a full handoff to the Autonomous Workforce, then it acts on that decision.
What Otto can actually do
Four capabilities carry most of Otto's weight today.
Conversational AI is the core interface. Employees submit requests in plain language, through chat, and Otto determines the intent, maps it to the right workflow or agent, and executes it end to end. A request like "I need access to the procurement system" does not get forwarded to a portal queue. Otto understands what is needed, checks the employee's role and permissions, routes the approval request, and confirms access once it is granted.
Enterprise Search retrieves answers across documents, wikis, databases, SharePoint, and internal knowledge sources, with results personalized to the employee's role, location, and department. The distinction from traditional search is that results are not a list of documents to read. Otto surfaces a direct answer, grounded in the specific context of the person asking.
AI Voice Agents handle requests through natural language conversations in multiple languages, without menu trees or hold queues. This has the most immediate relevance for service desks, contact centers, and any environment where employees or customers are not primarily working from a keyboard. The voice interaction goes through the same intent understanding and execution layer as text. The output is the same: completed work.
AI Data Explorer allows anyone to query enterprise data in plain language and receive analysis without involving an analyst or building a report. Questions like "Which departments had the highest volume of IT incidents last quarter and what were the most common categories?" get an immediate, structured answer. This capability makes data accessible to people who have always needed it but have never been able to retrieve it without help.
What ties all four together is memory. Context follows the person across channels, so a request started in chat during the morning and picked back up by voice in the afternoon does not need to be explained twice.
Every action runs through AI Control Tower
Speed without governance is a liability in most regulated industries, and ServiceNow built Otto around that constraint rather than in spite of it. Every routing decision and every workflow Otto triggers gets logged, checked against policy, and made explainable through AI Control Tower.
That governance layer expanded alongside Otto's launch. More than 30 new integrations now cover major clouds and enterprise systems including AWS, Azure, Google Cloud, SAP, and Workday. Runtime observability into how agents actually reason came through ServiceNow's acquisition of Traceloop, an AI observability company it acquired in March 2026. Identity governance, including the ability to detect a permission change on a model and automatically trigger a re-scoping workflow, runs through ServiceNow's acquisition of identity security firm Veza.
For a CIO or compliance lead evaluating Otto, this is the part worth paying closest attention to. Otto's autonomy is not sold separately from its audit trail. Every action it takes can be traced back to the policy, permission, or approval chain that allowed it in the first place.
Where Otto stands today
Otto is not everywhere in ServiceNow yet, and it is worth being precise about that before any procurement conversation starts. As of Knowledge 2026, Otto can be experienced first inside ServiceNow EmployeeWorks and AI Control Tower. ServiceNow has committed to extending it across the rest of its product portfolio through the remainder of 2026, building on the Australia release and the platform updates that follow it.
The commercial model changed alongside the product. On April 9, 2026, ServiceNow retired its five legacy licensing tiers and replaced them with three AI-native ones: Foundation, Advanced, and Prime. Otto's underlying components, Moveworks, Workflow Data Fabric, AI Control Tower, and process mining, are now bundled into every tier instead of sold as a separate line item. What changes as you move up the tiers is not whether AI shows up, but how much of a workflow it can complete without a person in the loop, moving from assisted work, to agentic multi-step automation, to fully autonomous completion through the Autonomous Workforce.
For organizations already licensed for Now Assist or building on the AI Platform, this matters immediately. Otto is less a new purchase decision than a new interface arriving on infrastructure many organizations are already paying for.
Who is already using Otto
The clearest proof point so far is ServiceNow EmployeeWorks, which put Otto's conversational capabilities in front of customers ahead of the wider rollout. In its first month, EmployeeWorks closed six deals, each exceeding $1 million in net new annual contract value, a result ServiceNow attributes directly to Otto completing work rather than only describing what to do next.
Early customers describe the shift in similar terms. At Siemens, conversational AI now delivers IT support and internal company communications straight to employees, cutting into the administrative back and forth that used to eat into their day. Honeywell's internal AI assistant, built on ServiceNow and Moveworks, has taken over the majority of its service desk conversations. Medtronic has built more than 100 custom use cases on the combined Moveworks and ServiceNow platform to help employees find what they need faster, so they can stay focused on patient-facing work.
“By leveraging ServiceNow, we've brought conversational AI directly to our employees, delivering seamless IT support and instant company communications. This eliminates administrative friction and preserves institutional knowledge, ensuring our workforce stays focused on what matters most: delivering exceptional outcomes for Siemens customers.”
— Elmar Spreitzer, Head of IT Digital Foundation, Siemens AG
None of these are hypothetical use cases pulled from a roadmap slide. They are production deployments at companies running the kind of complex, multi-system operations that make the completion problem worse in the first place.
Otto in the broader AI race
Otto is not the first attempt to make conversation the interface for enterprise software. Microsoft has Copilot, SAP has Joule, Salesforce has Agentforce, and each is chasing a version of the same idea: employees should be able to describe what they want rather than click through it screen by screen.
ServiceNow's case for Otto comes down to execution, not conversation. Any vendor can put a chat window in front of a large language model. Far fewer already own the approval chains, permissions, and cross-system workflows a request has to pass through to actually get done. ServiceNow's argument is that Otto works because the platform underneath it has been running those workflows for close to two decades. Whether that holds up is something buyers should test against their own workflows rather than take on faith.
The foundation Otto expects
Otto's routing decisions are only as good as the data, workflows, and permissions it is routing against. A platform with a fragmented data model, undocumented approval chains, or years of workflow customization built without governance in mind does not become coherent just because a conversational layer sits on top of it. Otto will surface whatever is underneath it, confidently, whether that foundation is ready or not.
That is the work organizations should be closing out now, well before Otto reaches every corner of their instance. It means a clean, validated CMDB and data model Otto can actually trust. It means workflows that are documented, owned, and consistent across business units instead of reinvented department by department. It means AI governance policies that exist before an agent is asked to act autonomously, not after something goes wrong.
FAQs
What is ServiceNow Otto?
Otto is ServiceNow's unified conversational AI experience, announced at Knowledge 2026. It combines Now Assist, Moveworks, and AI Experience into one interface that understands a request and carries it through to completion across ServiceNow's connected systems.
Is ServiceNow Otto the same as Now Assist?
No. Now Assist is the generative AI capability inside ServiceNow workflows, built for summarizing, drafting, and categorizing. Otto is the experience layer that decides which capability, Now Assist, Moveworks, or an autonomous agent, should handle a given request, then makes sure it actually gets done.
When is ServiceNow Otto available?
Otto is live today inside ServiceNow EmployeeWorks and AI Control Tower. ServiceNow has said it will extend Otto across its full product portfolio through the rest of 2026.
Does ServiceNow Otto cost extra?
Otto's underlying components now ship inside every Foundation, Advanced, and Prime tier as part of ServiceNow's April 2026 commercial model change, rather than as a separate add-on. What an organization pays largely comes down to tier and usage, not a standalone Otto license.
How is ServiceNow Otto governed?
Every action Otto takes runs through AI Control Tower, which logs interactions, enforces enterprise policy, and provides an audit trail and explainability for each decision Otto makes.
Do we need to prepare our ServiceNow instance before using Otto?
Yes. Otto routes requests based on the data, workflows, and permissions already configured in an instance. A fragmented data model or an undocumented workflow will produce unreliable results no matter how capable the conversational layer sitting on top of it is.
If you are evaluating ServiceNow, already running it, or trying to work out whether your instance is ready for what Otto expects underneath it, that is the conversation we have every day. Reach out to SYSUSA to talk through what an Otto-ready ServiceNow instance actually looks like for your organization.
