AI Agents

Enterprise

ServiceNow

Product overview

Name of Agent: AI Agents
Short description of agent: "ServiceNow AI Agents act autonomously to get work done" and "You can build custom AI agents using natural language in AI Agent Studio or deploy thousands of prebuilt AI agents" (link, archived)
Date of release: 12/03/2025 (link, archived)
Advertised use: "automate workflows, streamline operations and adapt to diverse business needs across IT, HR, CRM and more." (link, archived)
Monetisation/Usage price: None found
Who is using it?: enterprises
Category: Enterprise

Company & accountability

Developer: ServiceNow
Name of legal entity: ServiceNow, Inc.
Place of legal incorporation: Delaware (link, archived)
For profit company?: Yes
Parent company?: ServiceNow
Governance documents analysis: ToU (link, archived), AI AUP (link, archived)
AI safety/trust framework: Responsible AI White Paper (link, archived)
Compliance with existing standards: General ServiceNow compliance list, not agent specific (link, archived)

Technical capabilities & system architecture

Model specifications: You can select external models when creating AI agent (link, archived).
Observation space: Internet access, and company specific data (link, archived)
Action space: Edit data sources, return responses to the user (link, archived)
Memory architecture: System can "remember your preference or facts from previous interactions and use memories for more focused conversations" if setup to use memory. (link, archived).
User interface and interaction design: Form-like UI for designing agent: giving it prompts, setting execution mode, setting whether the agent displays outputs, etc.
User roles: Designer (user is designing an agent to deploy within ServiceNow's platform)
Component accessibility: Closed source

Autonomy & control

Autonomy level and planning depth: L1: user has full control over how to design the agent. Resulting agent is L4, agent seeks user approval to continue
User approval requirements for different decision types: Resulting agent can seek user approval before continuing
Execution monitoring, traces, and transparency: Agent shows some reasoning and data sources before providing answer
Emergency stop and shut down mechanisms and user control: Agent shows some reasoning and data sources before providing answer
Usage monitoring and statistics and patterns: AI agent decision logs panel on the right allows users to monitor activities of multiple agents

Ecosystem interaction

Identify to humans?: None found
Identifies technically?: - "ServiceNow/1.0.” is the user agent string for all ServiceNow instances and does not uniquely identify ServiceNow AI Agents (link, archived)
Interoperability standards and integrations: - MCP support, A2A (link, archived)
Web conduct: None found

Safety, evaluation & impact

Technical guardrails and safety measures: Unclear what guardrails are implemented in the background. The user can "define user access and Define data access. The former creates an ACL that determines who can discover or invoke the AI agent. The latter defines the data that the AI agent has access to once it’s invoked". (link, archived)
Sandboxing and containment approaches: None found
What types of risks were evaluated?: None found
(Internal) safety evaluations and results: None found
Third-party testing, audits, and red-teaming: None found
Benchmark performance and demonstrated capabilities: None found
Bug bounty programmes and vulnerability disclosure: Yes (link, archived)
Any known incidents?: AppOmni details agent-to-agent discovery / prompt-injection risk conditions for ServiceNow’s Now Assist AI-agent architecture (risk research, not an incident report) (link, archived)
← Back to 2025 Index