ServiceNow chooses Claude as default for enterprise AI automation
ServiceNow has named Anthropic’s Claude as the default model for its Build Agent and a preferred model across the ServiceNow AI Platform, signaling a shift from AI pilots to deeply embedded, production-grade automation.
Why platform-native AI matters now
Enterprises are moving beyond experiments and need AI that scales with security, governance, and measurable ROI, not just model benchmarks. ServiceNow already orchestrates tens of billions of workflows each year across IT, HR, customer service, and security. Embedding Claude into that fabric gives customers an on-ramp to agentic automation—systems that can reason over context, decide, and execute tasks—without stitching together point tools. For CIOs and CTOs, this is a signal that platform-native AI is maturing, with operational guardrails and integrations you can deploy at scale.
Claude as default across ServiceNow AI Platform
Claude becomes the default model for ServiceNow Build Agent, an AI-assisted builder for apps and automations that supports professional developers and citizen developers via natural language. ServiceNow expects Build Agent usage to grow several-fold over the next year as customers standardize on AI-native development practices. Claude is also a preferred model across the broader AI Platform, giving customers a governed path to advanced reasoning and code generation. Internally, ServiceNow is rolling out Claude and Claude Code to more than 29,000 employees, reporting large reductions in seller preparation time and faster engineering cycles.
From bolt-ons to AI-native, agentic workflows
The strategic change is not a new chatbot; it is the redesign of workflows so AI is the decisioning and execution layer. That means tighter access controls, usage monitoring, and auditability inside the ServiceNow platform, plus the ability to constrain actions to approved operations. The bet both companies are making is that outcomes improve when AI is woven into ticket triage, change management, and case handling, not used as an occasional add-on.
What customers get with ServiceNow + Claude
The deal spans development, deployment speed, and industry-specific solutions, with a focus on governed agentic automation.
Build Agent: Claude default for code and workflows
Build Agent now defaults to Claude for code generation and agentic workflow design, allowing teams to create apps and automations from natural language prompts. Developers can design, test, and execute autonomous tasks with visibility into actions taken and the ability to enforce guardrails. For organizations standardizing on citizen development, this lowers the barrier to deliver apps that previously required specialist resources.
Accelerated implementation and time-to-value
ServiceNow aims to cut implementation timelines for customers by roughly half by using Claude to accelerate configuration, adoption, and rollout. The approach also extends to partners who can use the same tooling to compress delivery cycles. For buyers, the value is less about model novelty and more about shaving weeks off deployment and getting to measurable outcomes sooner.
Regulated industry solutions on a governed AI platform
Early focus areas include healthcare and life sciences, where Claude will support research analysis and claims authorization inside ServiceNow’s governance model. The companies expect cycle times for approvals to move from days to hours while reducing cost-to-serve. Anthropic reports strong performance for Claude on medical and life sciences benchmarks, and the joint go-to-market will bundle these agentic patterns for regulated environments.
Operational impact, benchmarks, and ROI signals
Initial results highlight productivity gains and signal where enterprises should expect value to show up first.
Sales and engineering productivity gains
ServiceNow’s internal rollout shows up to a 95% reduction in seller preparation time by letting Claude synthesize account context, web intelligence, and enterprise data securely. Engineers are using Claude Code for code generation, review, debugging, and tooling, compressing the gap from idea to implementation. These are representative entry points for most enterprises: knowledge synthesis for go-to-market and AI-assisted software delivery.
Enterprise-grade AI guardrails and compliance
Embedding Claude within the ServiceNow AI Platform enables access control, usage monitoring, and compliance aligned to enterprise policies. Customers retain visibility into what the model does and can constrain actions to approved APIs and workflows. This architecture is key for regulated industries and global deployments where data residency, auditability, and human-in-the-loop approvals are nonnegotiable.
Economics and KPIs to track
The business case hinges on a few levers: reduced implementation timelines, higher developer velocity, and lower cost-to-serve in service operations. Track utilization of Build Agent, the share of workflows augmented by agentic automation, cycle-time reductions in ticket resolution or claims processing, and quality metrics such as error rates and rework.
What it means for telecom and global enterprises
For network operators, OEMs, and global enterprises, platform-native AI changes how OSS/BSS, customer care, and ITSM are delivered.
AI for network and IT operations
Agentic automations can triage incidents, recommend remediations, generate change requests, and execute standardized runbooks with approvals. For telcos, this supports zero-touch operations goals across NOCs and SOCs, from fault isolation to policy-safe configuration updates. Tight coupling with ServiceNow data models and CMDB improves context and reduces false positives.
AI for customer service and field operations
In customer care, Claude can draft responses, propose next-best actions, and automate back-office tasks like provisioning checks or billing adjustments with audit trails. For field service, it can generate work orders, summarize site histories, and validate completion against SLAs, accelerating mean time to repair while preserving compliance.
Governance, risk, and AI controls
Adopt a governance framework that aligns with standards such as the NIST AI Risk Management Framework and ISO/IEC guidance on AI management systems. Prioritize data-perimeter controls, prompt and action logging, human-in-the-loop for high-impact changes, retrieval-augmented grounding to enterprise knowledge, and rigorous red-teaming of agent actions before expanding scope.
What to watch in the next 6–12 months
Execution over the next 6–12 months will determine whether agentic automation scales from promising pilots to reliable production.
Agent reliability, tool-use, and action safety
Monitor how well Build Agent-created automations handle multi-step reasoning, ambiguous inputs, and rollback scenarios. Expect rapid iteration on tool-use policies, sandboxing, and approval workflows that keep humans in control while improving throughput.
Ecosystem openness and interoperability
ServiceNow says it will support an open ecosystem; watch for connectors, RAG patterns, and MLOps integrations that let customers mix models and keep data on platform. Alignment with industry frameworks (e.g., TM Forum Open APIs in telecom) will help translate agent decisions into compliant actions across OSS/BSS estates.
Measurable outcomes and adoption metrics
Track whether time-to-implement targets near 50% reductions hold across customer cohorts, whether Build Agent usage grows as projected, and whether claims of days-to-hours cycle times in regulated workflows generalize beyond early adopters.
Recommendations for CIOs and CTOs
Treat this as an opportunity to operationalize AI where work already happens, not as a separate tool.
Next 90 days: pilots and guardrails
Stand up a governed AI sandbox on ServiceNow; pilot Claude in three areas: IT incident triage, customer service case summarization and suggested actions, and developer productivity with Claude Code. Instrument baselines for cycle time, quality, and cost-to-serve. Establish action policies, approval gates, and audit logging before expanding agent privileges.
6–12 month roadmap: scale AI-native workflows
Refactor high-volume workflows to be AI-native, integrate retrieval from your knowledge bases, and expand to industry-specific use cases like claims or order fallout. Align governance with NIST AI RMF, define red/amber/green risk tiers for agent actions, and build a model evaluation pipeline to compare Claude with alternatives as requirements evolve. Focus on outcomes: faster deployments, lower operational cost, and higher CSAT, not just model scores.
