ServiceNow AI organization and tools
Summarize
Summary of ServiceNow AI organization and tools
The ServiceNow AI experience combines generative AI skills, agentic AI, and conversational engagement, all built on a foundation of quality data and a clear AI policy. These components work together to enable intuitive, efficient, and responsible AI-powered workflows and user interactions across the ServiceNow platform.
Show less
AI Policy and Governance
AI governance is established by your organization and executed using the AI Control Tower, which manages the full AI lifecycle—from intake through deployment and optimization—while ensuring compliance and ethical use. The AI Risk and Compliance application aligns with standards like the EU AI Act and NIST AI RMF to support responsible AI practices. The AI Guardian tool provides real-time monitoring and enforcement of guardrails against offensive content and security threats, configurable via the AI Admin Hub.
Key roles: AI Steward (data and governance), AI Asset/Product Owner (product delivery), AI Risk and Compliance Manager (risk mitigation), and Now Assist Admin (guardrail configuration).
Data Management
Successful AI outcomes depend on clean, structured data. ServiceNow offers tools like AI Data Kit for managing AI training data—enabling data curation, cleansing, synthetic data generation, and anonymization—and Knowledge Graph for contextualizing AI responses based on user roles and enterprise data. These tools help deliver personalized, accurate AI interactions.
Key roles: AI Data Kit Admin (dataset management) and Knowledge Graph Admin (schema design and auditing).
Content Search
ServiceNow Otto enhances AI Search by enabling discovery across internal content like knowledge articles, catalog items, and user data, as well as external repositories via External Content Connectors (e.g., Confluence, SharePoint). This broad content access improves AI-powered search relevance and user self-service capabilities.
Key roles: AI Search Admin (search configuration) and External Content Connectors Admin (external content management).
Generative AI Skills
ServiceNow Otto plugins offer pre-built, workflow-specific generative AI skills that automate tasks such as summarizing chats or generating resolution notes. The AI Admin Hub console allows admins to install, configure, monitor, and control these skills, including managing large language models and translation features. For customization, the AI Skill Kit enables building and testing tailored AI skills integrated with governance and data management tools.
Key roles: ServiceNow Otto Admin (plugin and skill management) and Skill Kit Admin (skill creation and publishing).
Agentic AI
Agentic AI uses autonomous agents capable of reasoning and multi-step task execution. AI Agent Studio allows building, testing, and managing these AI agents and workflows to automate complex processes like incident resolution or service requests. The Model Context Protocol (MCP) facilitates secure interactions between AI agents and ServiceNow platform features, enabling enhanced functionality.
Key roles: AI Agent Admin (agent/workflow management), MCP Client Admin, and MCP Server Administrator (MCP server management).
Conversational AI Engagement
ServiceNow Otto for Virtual Agent provides a natural language conversational interface across portals, mobile apps, and collaboration platforms (e.g., Microsoft Teams, Slack). It integrates AI agents, generative skills, catalog items, and knowledge sources to enable dynamic, context-aware conversations that guide users and automate tasks. Assistants configured via Assistant Designer deliver tailored experiences with customizable branding and channel deployment. The ServiceNow Otto panel offers contextual assistance within workflows.
Key roles: Virtual Agent Admin (assistant and topic management) and Now Assist Admin (Otto panel configuration).
What This Enables for ServiceNow Customers
- Implement responsible, governed AI aligned with organizational policies and regulations.
- Manage and prepare high-quality data to improve AI effectiveness.
- Enhance search and content discovery across internal and external repositories.
- Deploy and customize generative AI skills tailored to business workflows.
- Automate complex tasks with autonomous AI agents and orchestrated workflows.
- Engage users through natural, conversational AI across multiple channels and devices.
By leveraging these integrated tools and roles, ServiceNow customers can accelerate AI adoption, improve user self-service, reduce manual workload, and maintain compliance and security throughout their AI initiatives.
The ServiceNow AI experience includes generative AI skills, agentic AI, and conversational user engagement layer. A good AI experience depends on quality data and a coherent AI policy that functions as its guiding star. AI Search capabilities and other tools help connect the pieces.
The following infographic illustrates the components of the ServiceNow AI experience. You can think of each area as a layer, with AI policy as the basis for your implementation. Data plays an important role, since poor data will lead to poor outcomes, and the search experience drives both generative and agentic AI capabilities. The conversational engagement layer allows your users to communicate with AI and perform tasks and self-serve more easily.
AI policy
AI governance is defined by your organization and then implemented on the platform using AI Control Tower. AI Control Tower is designed to scale with your organization’s AI maturity. It enables end-to-end life cycle management of AI agents and models, from intake and evaluation to deployment and optimization. It also acts as a bridge between governance committees and implementation teams, ensuring that policy decisions are translated into technical configurations and operational controls. The platform includes the AI Risk and Compliance application, which comes with pre-built content aligned to the EU AI Act and NIST AI Risk Management Framework (AI RMF), helping organizations accelerate compliance and implement responsible AI practices from day one. For more information, see AI governance on the ServiceNow AI Platform.
For real-time guardrail enforcement for AI interactions, you can use AI Guardian to monitor and mitigate risks related to offensive content, sensitive topics, and security threats in generative AI outputs. Admins can configure these guardrails directly in the AI Admin Hub console, and logs are available to track detection events and guide decisions about blocking or escalation.
- AI steward [sn_ai_governance_ai_steward]
- The AI steward role manages the platform's data assets and AI capabilities, ensuring data quality, integrity, and ethical use. The steward also implements governance policies to maintain compliance, security, and responsible AI practices across ServiceNow applications and workflows.
- AI asset/product owner [ sn_ai_asset_mgmt.ai_asset_owner]
- The AI asset (or product) owner is responsible for driving the development and delivery of AI-powered products, ensuring alignment with business goals and user needs. They define product vision, manage feature prioritization, and translate business requirements into technical deliverables.
- AI Risk and Compliance Manager [sn_grc_ai_gov.ai_risk_and_compliance_manager]
- The AI Risk and Compliance Manager is tasked with identifying, assessing, and mitigating risks related to AI use within the enterprise. This role ensures AI compliance with laws, regulations, and internal standards while addressing issues such as bias, privacy, and transparency.
- Now Assist Admin [sn_nowassist_admin.nsa_admin]
- The ServiceNow Otto admin is responsible for enabling and managing the AI Guardian guardrails within the AI Admin Hub console. This includes configuring detection settings for offensiveness, prompt injection, and sensitive topics, as well as monitoring logs to assess harmful or malicious AI outputs. The admin can toggle guardrails on or off, choose actions like “Block and Log,” and tailor the configuration to match organizational risk tolerance and use cases.
Data management
You know the old saying, "Garbage in, garbage out?" Poor, incomplete, or unstructured data can lead to poor AI outcomes. To prepare your data for AI implementation, see Data readiness for implementing AI on the ServiceNow AI Platform.
ServiceNow Otto provides several tools to help you manage data and provide user context on your instance. The AI Data Kit helps organizations manage AI training data on their ServiceNow instance. It provides a centralized workspace to curate, generate, cleanse, and evaluate datasets, including importing data from tables, adding ground truth, and creating derived datasets. It supports synthetic data generation—up to 1,000 records at a time—and includes tools to anonymize sensitive information for privacy compliance. Integrated with the AI Skill Kit, it enables teams to test custom AI skills and measure performance.
Knowledge Graph makes AI in ServiceNow Otto more personal by using structured enterprise data to tailor responses to each user. It helps AI understand context, such as a user’s role, location, or assigned assets, making answers more relevant and accurate. It also supports intelligent slot-filling, automatically pulling in known information to streamline conversations and reduce repetitive questions. This results in faster, more personalized, and trustworthy AI interactions.
- ServiceNow Otto Data Kit admin [sn_data_kit.admin]
- This role is required to create, update, and publish datasets in AI Data Kit.
- Knowledge Graph admin [kg_admin]
- This role is required to design, manage, and audit Knowledge Graph schemas, including configuring nodes (tables), properties (columns), and relationships.
Internal and external content search
ServiceNow Otto in AI Search is the primary tool used to find content on your instance, whether it's in the form of knowledge articles, Service Catalog items, system users, Knowledge Graph schemas, or content in enhanced chats using ServiceNow® Otto for Virtual Agent.
ServiceNow Otto in AI Search is a plugin that enhances your existing AI Search configuration, adding support for ServiceNow Otto Genius Result answers.
You can expand the amount of content available to ServiceNow Otto in AI Search by using External Content Connectors. With this tool, AI Search applications can search content and metadata from external content repositories such as Atlassian Confluence Cloud and Microsoft SharePoint Online.
- AI Search admin [ais_admin]
- This role is required to manage configuration settings for the AI Search application.
- External Content Connectors admin [sn_ext_conn.xcc_admin]
- This role is required to manage configuration and crawl settings for the External Content Connectors application.
Generative AI skills
ServiceNow Otto product plugins provide workflow-specific, generative AI skills that are tailored to specific use cases. For example, a skill might summarize a customer chat, suggest next steps in a workflow, or generate resolution notes for a task. For a list of available skills by product, see Generative AI skills.
- Set up and configure skills to align with your business processes.
- Monitor and analyze usage, including adoption rates and performance metrics across the platform.
- Manage provider policies, data handling rules, and access controls.
- Turn on and configure the ServiceNow Otto panel.
- Manage large language models (LLMs).
- Configure translation for ServiceNow Otto.
If you need to customize or adapt existing skills, you can use the AI Skill Kit to build, customize, and test generative AI skills that are tailored to your organization’s workflows. With features like a prompt editor, visual builder, and evaluation tools, it helps you fine-tune performance and ensure quality. It also integrates with the AI Data Kit and AI Guardian for testing and governance, making it a flexible and secure way to extend AI across the ServiceNow platform.
- ServiceNow Otto Admin [sn_nowassist_admin.nsa_admin]
- This role is required to install ServiceNow Otto plugins, activate skills, and configure ServiceNow Otto preferences and settings.
- Skill Kit admin [sn_skill_builder.admin]
- This role is required to create, update, and publish skills in AI Skill Kit.
Agentic AI
In an agentic AI system, autonomous agents perform complex tasks by reasoning, planning, and executing actions across multiple steps. These agents are not just reactive; they can break down goals, make decisions, and coordinate with other agents or systems to complete workflows. In AI Agent Studio, AI agents and agentic workflows are built to handle tasks such as resolving incidents, managing requests, or navigating service catalogs using generative AI and orchestration logic. For a list of available agentic workflows by product, see Agentic workflows.
- Build or modify AI agents and agentic workflows using intuitive list views and guided setup tools.
- Simulate tasks and view execution logic to refine agent behavior.
- Access execution logs and version control to track performance and changes.
- Use the Settings page to enable AI Guardian for runtime safety, including detection of offensive content and prompt injection.
- Use the AI Agent Analytics dashboard to view usage stats, efficiency gains, and execution volume to help measure value.
The Model Context Protocol (MCP) client enables you to access the Model Context Protocol tools that are hosted externally and published using an MCP Server. MCP facilitates communication between an AI host application (like AI Agent Studio) and one or more MCP servers that expose specific capabilities such as tools.
The MCP Server Console is a tool for administrators to set up and manage Model Context Protocol (MCP) servers. These servers provide AI applications with access to ServiceNow features like incident lookups, case summaries, and workflow actions. The console lets you choose which capabilities to make available, ensuring everything stays secure and compliant. Then, from an MCP client, users can request information from the server and automate functionality using the available tools and data that the server returns.
- AI Agent Admin [sn_aia_admin]
- This role is required to create, manage, update, test, and delete AI agents and agentic workflows.
- MCP Client Admin [sn_mcp_client.admin]
- This role is required to administer the MCP client in AI Agent Studio. It is inherited from the AI Agent Admin [sn_aia.admin] role.
- MCP Server administrator [sn_mcp_server.admin]
- This role is required to administer the MCP Server Console.
Conversational AI engagement
ServiceNow® Otto for Virtual Agent acts as the conversational front door to your organization’s AI capabilities. It enables users to interact naturally with AI-powered assistants that understand intent, recommend actions, and deliver synthesized responses—all within familiar channels like the service portal, mobile app, and Microsoft Teams. This conversational layer connects users to a wide range of AI-powered resources, including AI agents, generative AI skills, catalog items, Virtual Agent topics, Q&A Genius results from AI Search, and even custom subflows and actions.
By integrating with the AI agent framework, Virtual Agent can discover and trigger agentic workflows that dynamically solve problems based on context. For example, a user might ask about ordering a device, and the system can invoke a multi-turn catalog ordering skill, refine the request, and complete the task, all in one conversation. Similarly, Q&A Genius capabilities allow users to ask open-ended questions and receive actionable answers, links, or follow-up options, improving deflection rates and reducing the need for live agent intervention.
ServiceNow® Otto for Virtual Agent also makes it easier to discover content without needing to know exact keywords. Users can search for and interact with catalog items, AI agents, and Virtual Agent topics using natural language. LLM topic discovery and Knowledge Graph slot-filling can return personalized responses and reduce repetitive questions.
To configure ServiceNow® Otto for Virtual Agent, you set up an assistant. An assistant is a container for a specific conversational experience that can be delivered in portals on your instance, mobile apps, Microsoft Teams, and Slack. Use the Assistant Designer to create, configure, and manage these assistants, as well as link them to conversational assets in Virtual Agent.
- AI agent enablement and orchestration
- Search sources
- Knowledge Graph schema usage
- Conversational assets, such as Virtual Agent topics, subflows, actions, generative AI skills, and AI agents
- The portals, channels, or apps the assistant will be used in
- Branding: the look and feel of the assistant
- Chat experience, such as greeting, closing, and fallback messages
You can also customize an assistant for the ServiceNow Otto panel. The ServiceNow Otto panel is available in workspace experiences and is typically used for contextual assistance within a specific workflow or record. ServiceNow Otto panel assistants have different configuration options and skill availability.
- Virtual Agent [virtual_agent_admin]
- This role, or the admin role, is required to install ServiceNow® Otto for Virtual Agent and set up and manage assistants in Assistant Designer. It is also required for users who create and manage Virtual Agent topics and assets.
- Now Assist Admin [sn_nowassist_admin.nsa_admin]
- This role is required to turn on and configure the ServiceNow Otto panel in the AI Admin Hub console.