What is ServiceNow Otto? The Complete Guide to ServiceNow's New AI Experience

ServiceNow Otto was announced at Knowledge 2026 as the next evolution of AI on the platform. It replaces Now Assist, unifies Moveworks, and runs on a new AI-native architecture. Here is what it actually is, how it works, and what it means for your instance.

If you have been following the ServiceNow AI story, you know the platform has gone through several phases — chatbots, Now Assist, AI Experience. Otto is the next step, and it is a significant one. It is not just a renamed product. It represents a fundamental change in how ServiceNow positions AI as the primary interface for enterprise work.

What is ServiceNow Otto?

ServiceNow Otto is a unified enterprise AI experience that combines three previously separate capabilities into a single product running on a new AI-native architecture:

  • Now Assist — ServiceNow's existing generative AI feature set (incident summarisation, code generation, search)
  • Moveworks — the enterprise AI assistant platform ServiceNow acquired, known for its conversational AI and cross-system capabilities
  • AI Experience — the unified AI interface announced at Knowledge 2025

The result is a single AI experience that employees interact with across every workflow, department, and system — without needing to know which tool, team, or application owns the underlying process.

How Does Otto Work?

Otto is built around a simple concept: turn intent into enterprise work. An employee says what they need — in natural language, by voice, or through search — and Otto figures out the rest. It understands the context of the request, identifies which systems and workflows are involved, routes the work, executes it, and completes it end to end.

Otto operates across four modes:

1. Conversational AI

The most familiar interface. Employees submit requests in natural language — "I need access to the Salesforce sandbox," "My laptop keyboard is broken," "I want to book a desk in the London office for Tuesday." Otto understands the intent, determines which workflow handles it, and executes or routes it without the employee navigating a service catalog or filling out a form.

Unlike Now Assist, which lived inside specific ServiceNow applications, Otto is designed to sit across the entire enterprise — aware of the employee's role, location, and existing access, and able to route across systems beyond ServiceNow.

2. Enterprise Search

Otto searches across documents, wikis, knowledge bases, databases, and enterprise applications with personalised results based on the user's role and context. This is the Moveworks capability integrated at the platform level — search that returns the right answer, not just the right document.

3. AI Voice Agents

Otto supports natural, multilingual voice interactions. This matters most for frontline workers, call centres, and service desk scenarios where typing is not the primary mode of interaction. Voice requests go through the same intent understanding and execution layer as text requests.

4. AI Data Explorer

Query enterprise data in plain language and generate actionable insights — without writing a report, building a dashboard, or involving an analyst. "How many P1 incidents were resolved in under four hours last quarter?" is a question any manager can ask and get an immediate answer.

What Does Otto Replace?

This is the question most ServiceNow practitioners are asking. The honest answer as of Knowledge 2026: Otto is the direction, not an immediate replacement.

Now Assist remains the current GA product. Otto is where Now Assist is heading — the brand and architecture will unify under Otto over the course of the Australia release and beyond. If you are deploying Now Assist today, you are building toward Otto. Your configurations, activations, and use cases will carry forward.

Moveworks as a standalone product will gradually be absorbed into Otto. If your organisation acquired Moveworks separately, the integration path will be through the Otto architecture.

How is Otto Different from Now Assist?

Now Assist was built feature by feature, application by application — summarisation for incidents, code generation in Studio, search in the portal. Each capability was activated separately and lived within its specific context.

Otto is architected differently. It sits above individual applications as a unified layer that understands the user, their context, and the full set of available workflows across the enterprise. The difference in practice:

  • Now Assist: "Use AI within this ServiceNow application"
  • Otto: "Use AI to complete work, and let Otto figure out which application handles it"

Otto also introduces multimodal interactions — text, voice, and data queries through a single interface — which Now Assist did not support.

What Does Otto Mean for ServiceNow Admins and Developers?

For admins, the immediate implication is understanding how Otto's governance model works. Otto executes work on behalf of users — which means the permissions, approvals, and guardrails you have set up in your instance directly shape what Otto can and cannot do. AI Control Tower is the governance layer that controls Otto's actions.

For developers, the architecture shift matters. Otto is built on the same Action Fabric that allows external AI agents to execute ServiceNow workflows. Understanding how intents map to flows, catalog items, and business rules — and how those are exposed through the platform's system of action — becomes increasingly important as Otto adoption grows.

For both, the practical preparation is the same: ensure your workflows, catalog items, and knowledge articles are clean, well-structured, and correctly permissioned. Otto's quality is entirely dependent on the quality of the underlying platform configuration. Garbage in, garbage out applies just as much to agentic AI as to any other automation.

When is Otto Available?

As of Knowledge 2026, Otto was announced and demonstrated in preview. The full Otto experience is rolling out through the Australia release and subsequent platform updates throughout 2026. Specific GA dates for each capability vary by feature — check the ServiceNow release notes for the Australia release for the current availability status.

Now Assist remains the current GA product for AI capabilities. If you are planning AI deployments for 2026, build on Now Assist with the understanding that the architecture and brand will evolve toward Otto.

Key Takeaways

  • Otto unifies Now Assist, Moveworks, and AI Experience into a single AI-native interface
  • It is designed to sit across the entire enterprise, not inside individual applications
  • It operates through conversational AI, enterprise search, voice agents, and data explorer
  • Now Assist remains the current GA product — Otto is the direction, not an immediate replacement
  • Platform quality (workflows, catalog, knowledge) directly determines Otto's effectiveness
  • AI Control Tower governs what Otto can do — permissions and guardrails matter more than ever

OTTO and what it means for how you build on ServiceNow

OTTO's multi-agent architecture requires a different design mindset than traditional Flow Designer automation. Flows are deterministic — you design the path and the platform follows it. OTTO agents are adaptive — they reason about the goal and choose their own path from available actions. This means developer work shifts toward: defining clear, well-documented action primitives (via Action Fabric), writing precise system prompts that constrain agent behaviour, building robust REST API endpoints that agents call, and configuring the governance rules in AI Control Tower that prevent agents from taking destructive or out-of-scope actions. The analogy: traditional automation is writing a script; OTTO-based automation is writing a job description and giving an employee the tools to do the job. The quality of outcomes depends on the quality of the instructions and boundaries, not on scripting every step.

OTTO activation and availability

As of 2026, OTTO is in early availability — available to select customers for pilot deployments rather than broadly GA. If you are interested in evaluating OTTO for your organisation, contact your ServiceNow account team for early access programme details. In the meantime, the best preparation is building the foundational layer: activating and configuring Now Assist, ensuring your REST API surface is clean and well-documented, reviewing your Flow Designer flows for agent-readiness (idempotent, well-parameterised, with clear success/failure outputs), and understanding the ACL architecture that governs what agents will be permitted to access.

How OTTO relates to the existing ServiceNow automation stack

ServiceNow has three layers of automation that developers need to understand as a stack. At the base: Business Rules — synchronous, event-driven, developer-coded logic that fires on specific table operations. Above that: Flow Designer — visual, process-oriented automation for multi-step workflows with human approvals and branching logic. At the top: OTTO — AI-orchestrated, goal-directed automation where the agent determines the path. Each layer has its appropriate use cases. Business Rules for immediate, deterministic data validation and computation. Flow Designer for defined business processes with known decision points. OTTO for complex, adaptive processes where the path depends on context, intermediate results, and reasoning about what action best achieves the goal. The three layers complement rather than replace each other — OTTO orchestrates flows and triggers business rules as part of achieving its goals, it does not replace them.

OTTO represents a genuine shift in how enterprise automation is designed and built. Practitioners who understand both the AI orchestration layer and the underlying platform — Flow Designer, REST APIs, ACLs — are positioned to lead OTTO implementations when they reach broad availability. Follow the early access programme through your account team and monitor releases through the Knowledge recap series.

Preparing your instance for OTTO

While OTTO is in limited preview, the preparation work is the same as for autonomous workforce broadly: clean CMDB data, well-structured Flow Designer flows with clear input/output contracts, comprehensive ACL security that supports fine-grained agent permissions, and robust notification workflows for human escalation. The practitioners who will be fastest to implement OTTO when it reaches GA are those who have already built these foundations — not because they anticipated OTTO specifically, but because these are the same foundations that support any sophisticated platform usage.

Already using Now Assist?

The NowSpectrum Now Assist Complete Practical Guide covers everything working professionals need to know — activation, real-world use cases, and what actually works in production.

Get the Now Assist Guide →
← Back to all posts