ServiceNow Otto Review 2026: What Changed, What Didn't, and Whether You Should Care
ServiceNow Otto consolidates Now Assist, Moveworks, and AI Experience into one AI layer. Here's what it actually changes for enterprise IT teams.
TLDR: ServiceNow Otto is a real consolidation — not just a rebrand. It merges Now Assist, the Moveworks acquisition, and AI Experience into a single conversational AI interface that works across chat, voice, mobile, and web. The Autonomous Workforce initiative and expanded AI Control Tower are genuinely new capabilities. But the hard truth from the earlier Now Assist era still applies: Otto amplifies whatever’s underneath it. Clean CMDB, structured knowledge bases, and mature ITIL processes become dramatically more capable. Messy foundations become sources of confident, autonomous mistakes. For existing ServiceNow customers with clean data, this is the upgrade worth pursuing at your next renewal. For everyone else, fix your data first.
Why This Matters Now
ServiceNow used Knowledge 2026 in Las Vegas to make the biggest enterprise AI announcement in the ITSM space this year. The headline: ServiceNow Otto replaces Now Assist as the primary AI experience across the entire platform. CEO Bill McDermott framed it as a shift from “platform of platforms” to “AI agent of agents.”
The timing matters for three reasons. First, ServiceNow processes over 100 billion transactions annually for its enterprise customers — any architectural shift at that scale ripples through IT budgets industry-wide. Second, “servicenow otto” is a brand-new keyword with near-zero search competition today, which means the first practitioner-level analysis will shape how IT buyers evaluate it. Third, the Autonomous Workforce initiative directly competes with Microsoft’s Agent 365 (GA May 1, 2026), Salesforce Agentforce, and IBM watsonx Orchestrate — all announced within the same two-week window.
If you’re a ServiceNow customer on renewal, an ITSM evaluator comparing platforms, or an IT leader trying to decode conference marketing from actual capability, this review is for you.
What ServiceNow Otto Actually Is
Otto is not Now Assist with a new logo. ServiceNow’s own product page makes this clear: “We’ve reimagined Now Assist, Moveworks, and AI Experience as ServiceNow Otto.”
That sentence deserves unpacking. Three distinct products — each with different architectures, user bases, and integration models — have been consolidated into one AI experience:
- Now Assist was ServiceNow’s native generative AI layer: incident summarization, KB drafting, resolution notes, virtual agent responses. It lived inside ServiceNow workflows.
- Moveworks was an independent employee-facing conversational AI platform that ServiceNow acquired. It excelled at cross-system task completion — resetting passwords, provisioning software, answering HR questions — by connecting to dozens of enterprise systems outside ServiceNow.
- AI Experience was ServiceNow’s experimental unified interface layer, previewed at Knowledge 2025.
Otto combines all three into a single interface where employees ask through chat, voice, mobile, or web, and the system routes requests across whatever backend systems are needed. Context persists across interactions — start a request on your phone, continue it on desktop, and Otto remembers what happened.
Earned insight: The Moveworks integration is the story most coverage is missing. Moveworks built deep connectors to 100+ enterprise systems before the acquisition — Okta, Workday, Jira, Slack, and dozens more. That cross-system reach is what makes Otto meaningfully different from Now Assist, which was limited to ServiceNow-native workflows. The question for existing customers is whether those Moveworks integrations survive consolidation intact, or whether they get simplified (read: weakened) to fit the ServiceNow data model. Ask your account team for a specific integration inventory before assuming everything carries over.
The Full AI Platform at a Glance
ServiceNow’s AI stack is now five interconnected layers. Understanding how they fit together is essential to evaluating Otto:
| Component | What It Does | What Changed at Knowledge 2026 |
|---|---|---|
| ServiceNow Otto | Unified conversational AI — chat, voice, mobile, web | NEW: Replaces Now Assist, Moveworks, and AI Experience |
| AI Agents | Autonomous task execution across ITSM, CRM, HR, SecOps | EXPANDED: AI Agent Studio for no-code/natural language agent building |
| AI Agent Orchestrator | Coordinates multi-agent teams for complex workflows | NEW: Multi-agent collaboration for cross-department processes |
| AI Agent Fabric | Connects third-party AI agents via A2A and MCP protocols | NEW: Agent2Agent protocol for cross-platform agent communication |
| AI Control Tower | Governance, compliance, observability, financial tracking | EXPANDED: 30+ integrations, identity governance (Veza partnership) |
What Works
The Unified Experience Is Genuinely Better
Now Assist’s biggest limitation was that it only worked inside ServiceNow. If an employee needed to reset an Okta password, provision a Jira project, and update their Workday profile, they were bouncing between three different systems. Otto — inheriting Moveworks’ cross-system architecture — handles all of these from a single interface.
The context persistence across channels is a meaningful improvement. ServiceNow claims that “every interaction picks up where the last one left off.” If that works as described — and early customer references from Honeywell, PayPal, Booking.com, and the NHL suggest it does for high-volume scenarios — it eliminates one of the most common employee frustrations with service desk interactions: repeating yourself.
AI Agent Fabric and the A2A Protocol
AI Agent Fabric is the most architecturally significant announcement at Knowledge 2026, and it got the least press attention. It embeds the Agent2Agent (A2A) protocol and Model Context Protocol (MCP) directly into the ServiceNow platform, allowing ServiceNow AI agents to communicate with third-party agents from Microsoft, Salesforce, SAP, and open-source frameworks.
This matters because real enterprise workflows cross system boundaries. An incident that starts in ServiceNow might require actions in AWS, updates in Salesforce, and notifications in Slack. Agent Fabric lets AI agents coordinate across those boundaries without custom integration code.
Tip: If you’re evaluating Agent Fabric, ask specifically about the A2A protocol’s maturity with your third-party stack. The protocol is new, and ServiceNow’s published integration count (30+ for AI Control Tower) likely represents monitored connections, not fully bidirectional A2A agent communication. Request a proof-of-concept with your actual vendor mix before committing.
AI Control Tower Expansion
The AI Control Tower now covers governance across ServiceNow-native and third-party AI agents — including real-time observability into agent behavior, automated compliance controls, and financial dashboards for tracking AI spend.
The Veza partnership for identity governance is notable: it extends oversight to human, machine, and AI agent identities simultaneously. In an era where agent sprawl is a genuine security risk (Gartner projects 40% of enterprises will experience a security incident from unauthorized AI agents by 2030), having unified identity governance across all three identity types is a real differentiator.
Where It Struggles
The “90% Resolution” Claim Needs Context
ServiceNow claims its own internal deployment handles over 90% of employee IT requests with cases resolved 99% faster compared to human agents. Those numbers come from ServiceNow’s internal IT organization — a company whose employees live and breathe the platform daily, whose knowledge base is impeccable, and whose CMDB is presumably best-in-class.
Your environment is not ServiceNow’s internal IT. Independent analysis of Now Assist deployments (Otto’s predecessor) consistently showed autonomous resolution rates of 10-20% without human escalation in typical enterprise environments. That gap between internal showcase numbers and real-world deployment is the single biggest thing to calibrate when evaluating Otto.
Warning: Conference keynote numbers are always internal best-case scenarios. When ServiceNow says “90% of IT requests handled by AI,” that’s their own IT team using their own platform with their own data. Ask for customer reference resolution rates from organizations with similar CMDB maturity, ticket volume, and knowledge base quality to yours. If ServiceNow can’t provide them for Otto specifically (it just launched), benchmark against Now Assist deployments — the underlying AI hasn’t fundamentally changed.
Autonomous Workforce Is a Vision, Not a Deliverable
The Autonomous Workforce initiative — AI specialists that do “jobs, not just tasks” across IT, CRM, employee experience, and security — is the most ambitious claim from Knowledge 2026. ServiceNow positions these as autonomous agents that orchestrate teams of other AI agents to deliver outcomes from start to finish.
The reality: this is a roadmap item dressed as a product launch. The AI Agent Studio for building custom agents is real. The AI Agent Orchestrator for coordinating multi-agent teams is real. But “AI specialists that replace human roles” is a trajectory, not something you can deploy on Monday. Enterprises that treated Now Assist’s agentic AI promises literally in 2025 and expected autonomous Tier 1 service desks were disappointed. Apply the same skepticism to Autonomous Workforce timelines.
Migration Complexity for Existing Now Assist Customers
If you’ve customized Now Assist — and most enterprise customers have — the migration to Otto is not a toggle. ServiceNow’s FAQ carefully states that Otto is “a fundamental shift in the experience” built on “a new AI-native architecture.” Translation: your existing Now Assist customizations, trained models, and workflow integrations may need rework.
ServiceNow has not published a detailed migration guide as of May 2026. Until that exists, budget for migration effort the same way you’d budget for a major platform upgrade — not a feature release.
ServiceNow Otto Strengths:
- Genuine consolidation of three products (Now Assist, Moveworks, AI Experience) into one interface
- Cross-system task completion inherited from Moveworks — not limited to ServiceNow-native workflows
- Context persistence across chat, voice, mobile, and web channels
- AI Agent Fabric with A2A and MCP protocols for third-party agent interoperability
- AI Control Tower with identity governance covering human, machine, and AI agent identities
- AI Agent Studio enables natural-language agent building without code
ServiceNow Otto Weaknesses:
- Autonomous resolution rates in real deployments are far below ServiceNow’s internal showcase numbers
- Autonomous Workforce is a roadmap vision, not a deployable product today
- Migration path from customized Now Assist deployments is undefined
- CMDB accuracy requirements haven’t changed — dirty data still equals confident AI mistakes
- New architecture means existing Now Assist customizations may need rework
- Pricing details for Otto-specific features remain unpublished
Otto vs. the Competition
Every major enterprise platform vendor made an AI agent announcement in the same two-week window. Here’s how they compare:
| Dimension | ServiceNow Otto | Microsoft Agent 365 | Salesforce Agentforce | IBM watsonx Orchestrate |
|---|---|---|---|---|
| Primary function | Unified AI experience for task completion | Agent governance and control plane | CRM-native autonomous agents | AI workflow orchestration |
| What it does | Handles employee requests across systems via conversation | Discovers, inventories, and governs all AI agents | Automates sales, service, and marketing workflows | Orchestrates multi-step AI processes for business users |
| Best for | ServiceNow-heavy enterprises needing cross-system AI | M365 shops managing agent sprawl | Salesforce customers automating CRM workflows | IBM/hybrid-cloud enterprises with complex orchestration |
| Agent interop | AI Agent Fabric (A2A + MCP) | Entra Agent ID + registry across platforms | Einstein Trust Layer + MuleSoft integrations | Open-source compatible, multi-model |
| Governance | AI Control Tower (30+ integrations) | Native Defender/Purview/Entra integration | Einstein Trust Layer, audit trails | watsonx.governance |
| Pricing model | Bundled into platform tiers + consumption | $15/user/mo standalone or M365 E7 ($99/user/mo) | Consumption-based ($2/conversation) | Enterprise negotiated |
| Maturity | Week 1 (Knowledge 2026 launch) | GA May 1, 2026 | Production since late 2024 | Production since 2023 |
Earned insight: The real competitive divide isn’t features — it’s where your data lives. Otto is strongest when your workflows already run through ServiceNow and you need AI to span beyond it. Agent 365 is strongest when you need to govern agents regardless of where they run. Agentforce is strongest when your revenue workflows live in Salesforce. None of these products compete head-to-head; they compete for which vendor becomes the “system of record” for your AI agents. That’s the strategic bet you’re making, not a feature checklist.
Pricing Reality
ServiceNow restructured its AI pricing in April 2026, collapsing Now Assist add-on charges into three bundled tiers:
| Tier | AI Capabilities Included | Typical Range |
|---|---|---|
| Foundation | Generative AI tasks: summarization, data extraction, insights | $100-130/fulfiller/month (estimated) |
| Advanced | Deterministic + AI agent-executed workflows | $130-160/fulfiller/month (estimated) |
| Prime | Full autonomous role replacement, Autonomous Workforce features | Not published — enterprise negotiated |
ServiceNow does not publish list prices. These ranges come from contract analysis of existing deployments and should be treated as estimates, not quotes. All tiers include some level of “Assist unit” consumption for generative AI actions; overages are billed separately.
The real cost picture for a 50-fulfiller deployment:
| Cost Layer | Year 1 Estimate | Notes |
|---|---|---|
| Platform licenses | $78,000 - $96,000 | Advanced tier, 50 fulfillers |
| Assist unit overage | $10,000 - $30,000 | Depends on AI adoption rate |
| Implementation partner | $150,000 - $250,000 | SI fees for configuration, migration, training |
| Internal admin | $80,000 - $120,000 | 1-2 FTE ServiceNow admins |
| CMDB cleanup | $30,000 - $60,000 | Often overlooked, always necessary |
| Total Year 1 | $348,000 - $556,000 | Before any Autonomous Workforce premium |
Pricing research verified May 6, 2026 from ServiceNow product pages and existing contract benchmarks. ServiceNow does not publish list prices; figures are practitioner estimates.
Warning: Otto’s pricing relative to standalone Now Assist hasn’t been clarified. If Otto replaces Now Assist at no additional cost within existing tiers, this is straightforward. If Otto’s cross-system capabilities (inherited from Moveworks) require a separate SKU or higher tier, your renewal could surprise you. Get written confirmation from your account team on what Otto access costs under your current contract before Knowledge 2026 excitement fades.
Who Should Upgrade to Otto Now
Good fit:
- Existing ServiceNow customers on Foundation or Advanced tiers who are already using Now Assist and want cross-system task completion
- Organizations running Moveworks alongside ServiceNow who will benefit from a single consolidated interface
- Enterprises with 5,000+ employees where service desk volume justifies AI investment
- IT teams with clean CMDBs (85%+ discovery accuracy) and structured knowledge bases
- Organizations already using AI Control Tower who want expanded governance across third-party agents
Not a good fit (wait or skip):
- Companies with poor CMDB hygiene — Otto will make confident mistakes faster, not fewer
- Mid-market organizations under 500 employees where the cost/complexity ratio doesn’t justify ServiceNow
- Teams that heavily customized Now Assist workflows — migration path is unclear and migration cost is real
- Organizations evaluating ServiceNow for the first time — start with the base platform, prove value, then layer in AI
- Anyone expecting “Autonomous Workforce” to replace human agents in 2026 — that’s a 2027-2028 story at best
Bottom Line
ServiceNow Otto is a genuine consolidation, not a conference-season rebrand. Merging Now Assist, Moveworks, and AI Experience into a single conversational AI interface with cross-system reach is a meaningful architectural step forward. AI Agent Fabric’s A2A and MCP protocol support positions ServiceNow as a legitimate “agent of agents” orchestrator — not just an ITSM platform with AI bolted on.
But the fundamental equation hasn’t changed. Otto amplifies whatever’s underneath it. Organizations with mature ServiceNow deployments, clean data, and documented processes will get dramatically more value from Otto than they got from Now Assist. Organizations with CMDB debt, fragmented knowledge bases, and ad-hoc processes will find that Otto automates their dysfunction more efficiently.
The Autonomous Workforce vision is ambitious and directionally correct. But treat it as a 2027+ deliverable, not a 2026 deployment. The productized, deployable value today is in Otto’s unified conversational interface, Agent Fabric’s interoperability, and AI Control Tower’s expanded governance. Start there.
For existing ServiceNow customers with clean platforms: evaluate Otto seriously at your next renewal. The cross-system capabilities alone (thank the Moveworks acquisition) fill a gap that Now Assist never addressed. For everyone else: fix your data first. No AI experience — no matter how well-consolidated — can compensate for a messy foundation.
Rating: 4.2 / 5 for mature ServiceNow customers with clean data. 2.5 / 5 for organizations starting fresh or carrying significant CMDB debt.
Related Articles
- ServiceNow AI Review 2026: Now Assist, Agentic AI, and What Practitioners Actually Think
- Microsoft Agent 365: The Enterprise Control Plane for AI Agents
- Best AIOps Platforms for Mid-Market Companies (2026)
- AI Governance Framework for IT Teams
- PagerDuty vs Opsgenie vs ServiceNow ITSM: 2026 Comparison
Discussion