Now Assist is ServiceNow's generative AI product, embedded directly into the platform across ITSM, HR, CSM, Field Service, Security Ops, and Creator. Rather than a separate AI tool, it appears at the point of work — inside the incident form, the case view, Studio, and Flow Designer — so agents and developers get AI assistance without switching context.

The four core tasks Now Assist performs:

Summarisation — Condenses long incident work notes, case history, or chat logs into a brief so agents understand context in seconds.

Content generation — Drafts email replies, resolution notes, and knowledge articles based on the record context.

Code and flow generation — Writes ServiceNow scripts and Flow Designer actions from natural language descriptions in Studio.

Search — Returns a direct generated answer to a natural language query rather than a list of article links.

They serve different audiences and are designed to work together, not compete.

Virtual Agent is a conversational chatbot for end users (employees or customers). It follows Topic Designer conversation flows, asks questions, and takes actions like resetting a password or checking ticket status. It can deliver responses using generative AI for more natural language.

Now Assist is an AI layer embedded inside ServiceNow for human agents, developers, and admins. It does not have a conversation flow — it generates content in context: summarises a ticket the agent just opened, drafts a reply the agent is composing, or writes a script the developer is building.

Virtual Agent deflects L0 and L1 tickets by resolving the user's request autonomously. Tickets that still reach a human agent are handled faster because Now Assist provides instant context, drafts replies, and generates resolution notes.

The clearest way to remember it: Virtual Agent talks to the end user. Now Assist helps the human agent doing the work.

Predictive Intelligence (PI) and Now Assist are both AI features in ServiceNow, but they use different technology and serve different purposes.

Predictive Intelligence uses classic machine learning trained on your own historical data to predict structured field values: which category an incident belongs to, which assignment group should handle it, which knowledge articles are similar. It predicts from patterns in your own instance data. No LLM involved.

Now Assist uses large language models (LLMs) to generate new content — summaries, draft replies, scripts, flow logic. It is generative AI, not predictive AI.

In practice they work together: Predictive Intelligence categorises and routes the ticket automatically at creation. Now Assist helps the agent respond faster once the ticket is assigned. Different technology, different lifecycle stage, different output.

AI Search replaced the legacy Zing search engine. The fundamental difference is the output: standard search returns a ranked list of matching records; AI Search returns a direct generated answer.

Standard search returns a list of articles and records ranked by keyword relevance. The user must open and read several articles to find the answer.

AI Search uses semantic search (Natural Language Understanding) to match intent, not just keywords. Combined with Now Assist, it generates a 2–4 sentence answer at the top of results, synthesised from the most relevant articles, with citations to the source content.

Example: A user searches “I can't get into my computer.” Standard search returns articles with those exact words. AI Search understands the intent and surfaces password reset and MFA lockout articles, then generates a direct answer showing the resolution steps.

AI Search also self-improves — it learns from which results users click and adjusts relevance rankings automatically. The downstream impact: users resolve more issues at the search step without submitting a ticket.

AI Agents in ServiceNow are autonomous software workers that can sense context, make decisions, execute actions across the platform, and adapt based on results — with minimal human intervention. They were introduced to general availability in the Yokohama release (Q1 2025).

Unlike a Virtual Agent that follows a scripted conversation flow, an AI Agent is given a goal and determines its own steps at runtime. It reads the current situation, decides the best action, executes it, checks the result, and continues until the goal is achieved or escalation is needed.

Examples of what OOTB AI Agents handle autonomously:
- Password resets and MFA issues (L1 IT Specialist)
- Access provisioning requests
- HR onboarding task coordination
- Security alert triage and initial containment

The key distinction from traditional automation: flows follow pre-authored steps. AI Agents determine their own steps based on the current context and goal.

Agentic AI refers to AI systems that can take sequences of actions autonomously to complete complex, multi-step goals — without requiring a human to direct each step. The defining characteristic is that the agent plans, decides, acts, and adapts based on the current state of the work, not a pre-scripted path.

Previous AI features in ServiceNow were reactive: Now Assist responded when an agent opened a form. Predictive Intelligence predicted when a ticket was created. These features assist a human at a specific moment.

Agentic AI is proactive: it can receive a goal and work through a multi-step resolution independently. “Resolve this incident” is a valid instruction. The agent diagnoses, acts, verifies, and closes — without needing a human to approve each step (unless configured to require it).

The term “agentic” comes from the word agency: the capacity to act independently toward a goal.

The AI Agent Orchestrator is the system that coordinates multiple AI Agents working in parallel or in sequence to complete complex, cross-domain tasks. It acts as the manager layer above individual specialist agents.

When a high-level goal arrives (e.g., complete employee offboarding for user X), the Orchestrator:
1. Breaks the goal into sub-tasks
2. Assigns each sub-task to the appropriate specialist agent
3. Manages dependencies (runs independent tasks in parallel; waits for prerequisites before starting dependent tasks)
4. Handles exceptions if an agent fails or hits an unexpected state
5. Synthesises results when all agents confirm completion

Example: An offboarding request triggers the Orchestrator. It coordinates: ITSM agent (revoke access), HR agent (process final payroll), Facilities agent (reassign workspace), Security agent (invalidate SSO tokens) — all running in parallel, completing in minutes.

AI Control Tower is ServiceNow's centralised platform for governing all AI deployed across the enterprise — including AI systems outside ServiceNow. It has five core capabilities:

Discover — Automatically finds all AI assets deployed across the enterprise, including AI in AWS, Azure, GCP, SAP, Salesforce, and 30+ other systems. Creates a unified AI asset registry.

Observe — Real-time monitoring of AI behaviour at runtime. Detects when agent decisions drift from expected patterns. Powered by Traceloop (acquired March 2026).

Govern — Maintains AI asset ownership, risk classification, and maps AI deployments to regulatory frameworks (GDPR, HIPAA, SOX). Produces audit-ready compliance reports.

Secure — Governs AI agent identities and access rights via Veza (acquired March 2026). Applies least-privilege to AI agents and can shut down an agent operating outside its permitted scope in real time.

Measure — Cost tracking and ROI dashboards showing spend per model, per use case, and efficiency benefits achieved.

ServiceNow's Autonomous Workforce is the suite of pre-built AI specialist agents for specific business functions. These agents come pre-trained and pre-configured for their domain.

L1 IT Service Desk Specialist — The first released (Yokohama, Q1 2025). Handles password resets, access provisioning, MFA issues, network troubleshooting. Resolves L1 tickets from creation to closure autonomously.

HR Service Delivery Specialist — Benefits enquiries, leave requests, onboarding coordination, HR policy questions.

Customer Service Specialist — Order status, account changes, complaint triage, escalation routing in CSM.

Security Operations Specialist — Alert triage, threat intelligence enrichment, initial containment steps.

Finance and Procurement Specialist — Invoice matching, PO validation, approval routing, exception handling.

Each specialist learns from outcomes over time, improving accuracy and resolution rate with each completed task. Additional roles are rolling out through 2026.

ServiceNow provides three distinct data protection options that organisations choose from based on their compliance requirements.

Now LLM (on-platform) — Data is processed entirely within ServiceNow infrastructure. No external API call is made. No data passes to any third party, even temporarily. This is the option for organisations where data must never leave the platform under any circumstances.

Partner LLMs with zero data retention — When using Azure OpenAI, Anthropic Claude, or Google Gemini, ServiceNow negotiates zero data retention agreements: the provider processes the prompt but does not store, log, or train on the data.

PII masking before external calls — Fields configured as sensitive (names, phone numbers, email addresses) are anonymised in the prompt before it is sent to any external LLM. The response is de-anonymised before being shown to the agent.

Additionally, tenant isolation ensures no data from one ServiceNow customer is ever used as context for another customer.

The Now Assist Skill Kit (NASK) is ServiceNow's low-code framework for building custom AI skills beyond the out-of-the-box Now Assist capabilities. It allows developers and admins to extend the AI platform to address organisation-specific needs.

A custom skill built with NASK can:
- Query a specific knowledge base, table, or external API as a context source (via Retrievers)
- Run server-side logic to pre-process retrieved data (via Scripts)
- Trigger existing Flow Designer subflows as part of the AI response pipeline (via Subflows)
- Chain multiple skills together, where one skill's output becomes the next skill's input (Skill Chaining)
- Use a custom LLM model (BYOLLM) instead of the platform default

NASK is the bridge between “Now Assist does what the platform ships” and “Now Assist does what our organisation specifically needs.”

ServiceNow completed the acquisition of Moveworks in December 2025 for approximately $2.85 billion. Moveworks was an enterprise AI assistant platform deployed at hundreds of large organisations.

What Moveworks brought to ServiceNow:

Conversational AI front-end — A natural language interface for employees to resolve IT and HR requests through Teams, Slack, or email — more polished and conversational than a standard Virtual Agent.

Enterprise search across systems — Ability to find answers across company systems simultaneously: SharePoint, Confluence, Jira, and ServiceNow itself, in one query.

Autonomous task completion — End-to-end handling of password resets, access requests, and HR enquiries without human agent involvement.

The combination: Moveworks brought the front-end conversational experience. ServiceNow brings the back-end workflow engine, governance, and platform depth. Together they cover the full arc from employee question to autonomous resolution.