ServiceNow Brazil Release - 8 new AI features (AI Assistance -->an Autonomous Enterprise Workforce)

[John Zhang1]

 

Introduction: A New Era of the ServiceNow AI Platform

The ServiceNow Brazil release represents another major milestone in the evolution of the ServiceNow AI Platform. While previous releases focused on introducing generative AI, enhancing Now Assist, and helping developers work more efficiently with AI, Brazil takes the next significant step: moving from AI-assisted work to AI-driven autonomous execution.

If the Australia release helped organizations adopt AI-assisted development and automation, Brazil takes this evolution further by introducing more autonomous capabilities across application development, enterprise workflows, testing, troubleshooting, agent orchestration, and governance.

The release extends AI beyond generating scripts and configurations. AI can now help understand business requirements, develop applications, generate and execute tests, troubleshoot problems, coordinate workflows, and prepare solutions for deployment—with human oversight and enterprise governance.

Another notable change is the rebranding of Now Assist to ServiceNow Otto, reflecting the broader evolution of ServiceNow's AI capabilities.

For ServiceNow architects, developers, platform owners, and enterprise leaders, Brazil creates an opportunity to rethink how we design, implement, operate, and govern enterprise platforms.

Early Access began on September 24, 2026, with General Availability listed for Q4 2026.

Let's explore the most important AI innovations and what they mean for ServiceNow professionals.

1. Autonomous Engineer: From Requirements to Implementation

One of the most exciting innovations in the Brazil release is Autonomous Engineer, which takes AI-assisted application development to another level.

Traditionally, ServiceNow developers spend considerable time analyzing requirements, designing technical solutions, creating tables, writing scripts, configuring workflows, developing integrations, and building test cases.

Autonomous Engineer introduces a specification-driven approach that helps transform business and technical requirements into implementation activities.

Its capabilities include:

• Requirement analysis: Understand written requirements and uploaded documents.

• Implementation planning: Break requirements into development tasks with acceptance criteria and test conditions.

• Application development: Generate and implement approved development tasks.

• Automated testing: Generate and execute Automated Test Framework (ATF) tests.

• Self-healing: Attempt to resolve test failures and rerun tests.

• Deployment preparation: Package completed work into update sets for further review and deployment.

An important distinction is that developers retain control of the implementation process. They can review and approve the implementation plan before execution, while unresolved issues are surfaced for human review.

Real-world use case:

Imagine an organization needs a custom ServiceNow application to manage IT service requests.

Instead of manually creating tables, forms, Business Rules, ACLs, Flow Designer workflows, and test cases, the development team provides a detailed technical specification to Autonomous Engineer.

The AI can translate the requirements into an implementation plan, build approved components, generate tests, and prepare the changes for User Acceptance Testing (UAT).

Architectural value:

This capability allows architects and senior developers to spend more time on solution architecture, data modeling, security, integrations, and platform governance rather than repetitive development activities.

2. Build Agent: AI-Powered Application Development

Build Agent brings conversational AI directly into the ServiceNow application development lifecycle.

Developers can describe what they want to build using natural language rather than starting with scripts and configurations.

Key capabilities include:

• Creating and modifying full-stack applications.

• Generating server-side and client-side components.

• Managing application files and user interface components.

• Explaining existing application code.

• Generating technical documentation.

• Creating custom AI agents, skills, and agentic workflows.

• Generating and executing ATF test suites.

• Performing quality checks and packaging development changes.

Another important improvement is the convergence of ServiceNow Studio and ServiceNow IDE, bringing development activities into a more unified experience.

Why this matters:

Traditionally, developers spend significant time understanding existing applications, navigating scripts, troubleshooting Business Rules, and maintaining technical documentation.

Build Agent can reduce this effort and accelerate application development while allowing developers to review and control generated implementations.

However, AI-generated code must still follow enterprise development standards, security requirements, coding practices, and platform governance policies.

3. AI Agent Studio: A Unified Experience for Building and Managing AI Agents

AI Agent Studio continues to evolve as the central environment for creating, configuring, managing, and deploying AI agents.

Brazil introduces a redesigned experience with additional capabilities for agent development, operational management, security, and performance monitoring.

Important enhancements include:

• A redesigned interface for creating and managing AI agents.

• Improved agent setup and deployment processes.

• Better AI agent version management.

• Built-in automated evaluations.

• Improved visibility into agent access and security.

• More structured management of agentic AI assets.

• Issue tracing and suggested optimizations on agent results pages.

• The ability to add or remove agents and tools from built-in AI agents.

One particularly important improvement is the integration of automated evaluations into the agent development lifecycle.

Organizations need to evaluate whether an AI agent:

• Produces accurate and relevant responses.

• Follows business policies.

• Invokes the correct tools.

• Respects security and access requirements.

• Completes its assigned tasks successfully.

Brazil also introduces additional operational controls, including detection and disabling of runaway AI agent triggers to prevent unintended consumption.

For enterprise administrators, deny-by-default ACLs for specified agentic AI record types provide another important security layer.

Architectural value:

AI Agent Studio is becoming more than an agent-building environment. It is evolving into an important part of enterprise AI development, testing, security, and lifecycle management.

4. Test Agent: Intelligent Testing and Self-Healing

Testing is traditionally one of the most time-consuming activities in ServiceNow implementations.

Brazil introduces AI-powered testing capabilities through Test Agent to help automate the testing lifecycle.

Key capabilities include:

• Generate ATF tests from natural-language prompts.

• Execute generated tests.

• Analyze test failures.

• Help troubleshoot application issues.

• Attempt to repair failing tests.

• Update tests based on testing results.

Real-world use case:

Consider a developer modifying an Incident Management workflow.

Test Agent can help generate test scenarios to validate:

1. Incident creation.

2. Incident assignment.

3. Priority determination.

4. Approval processing.

5. Incident resolution and closure.

If a test fails, AI can analyze the failure, investigate potential causes, and assist with troubleshooting.

Architectural value:

This capability can improve regression testing efficiency, accelerate development cycles, and reduce the manual effort required to maintain automated test coverage.

However, generated tests should always be reviewed to ensure that they validate actual business requirements rather than simply confirming that the application executes without errors.

5. AI Agent Advisor: Discover Automation Opportunities

Another important innovation is AI Agent Advisor.

Instead of requiring organizations to manually identify every opportunity for AI automation, AI Agent Advisor analyzes actual instance data to recommend potential automation opportunities.

Its capabilities include:

• Analyzing instance data to identify automation opportunities.

• Recommending automations with potential efficiency improvements.

• Matching opportunities with existing AI agents.

• Identifying opportunities that require new AI agents.

• Supporting the creation and deployment of additional agents.

Automation opportunities are also surfaced through the AI Agent Studio home page.

Why this matters:

Organizations frequently struggle to identify where AI agents can deliver measurable business value.

AI Agent Advisor helps shift the conversation from simply deploying AI technology to identifying specific business processes where AI can reduce manual effort and improve operational efficiency.

For platform owners, this provides a more practical starting point for building an enterprise AI adoption roadmap.

6. Autonomous Workforce and AI Specialists

Brazil further expands the concept of an autonomous workforce through AI specialists.

AI specialists are designed to autonomously resolve high-volume, low-effort tasks while handing off activities outside their scope to human employees.

Important capabilities include:

• A unified persona across different channels.

• Deployment across multiple teams.

• Team-specific customization.

• Autonomous execution of repeatable tasks.

• Human intervention when exceptions or complex scenarios arise.

The L1 IT Service Desk AI Specialist is an important example.

It can help automate high-volume, repeatable Level 1 IT incidents while allowing human service desk agents to intervene when necessary.

Real-world use case:

An employee reports an issue accessing a business application.

An AI specialist can understand the request, gather relevant information, execute approved troubleshooting actions, and attempt to resolve the incident.

If the issue requires additional permissions, complex investigation, or specialized technical support, it can escalate the incident to the appropriate human team.

Business value:

This approach can help organizations improve service desk productivity, reduce repetitive manual activities, and allow employees to focus on more complex support activities.

7. Model Context Protocol (MCP😞 Connecting AI Agents to Enterprise Tools

One of the most strategic developments in Brazil is the continued expansion of Model Context Protocol (MCP) capabilities.

MCP provides a standardized approach for AI agents to discover and interact with external tools and services.

This creates opportunities for ServiceNow AI agents to interact with other enterprise platforms and external AI ecosystems.

Brazil introduces several important MCP enhancements.

MCP Server Console

The MCP Server Console provides improved administration and configuration capabilities, including:

• OAuth client registration within the configuration wizard.

• Tool annotations to help compatible clients, such as Claude, apply appropriate permission policies.

• Support for creating MCP applications with interactive interfaces.

Enterprise Identity Integration

Organizations can register external authorization servers, such as Microsoft Entra ID and Okta, as trusted token issuers for the MCP Server.

This helps organizations align external AI integrations with existing enterprise identity and authentication strategies.

Product-Specific MCP Servers

Brazil expands MCP support across different ServiceNow product areas.

Examples include:

• ITSM MCP Server for incidents, change requests, and on-call scheduling.

• ITOM MCP Server.

• Strategic Portfolio Management MCP Server.

• Contract Management Pro MCP Server.

• Retail MCP Server.

Why MCP Matters

Enterprise AI becomes more powerful when agents can access enterprise data and execute approved actions across multiple systems.

For example, an IT Operations AI agent could coordinate activities across ServiceNow, infrastructure monitoring platforms, cloud services, and approved external enterprise systems.

However, MCP connections must be carefully governed.

Organizations should establish controls for:

• Tool permissions and authorization.

• Credentials and authentication.

• Data access and privacy.

• External system connectivity.

• Auditability and monitoring.

The objective is to create secure, controlled connections between ServiceNow AI agents and the broader enterprise technology ecosystem.

8. Intelligent Approvals: Bringing AI into Business Decisions

Another interesting innovation in Brazil is Intelligent Approvals.

Traditional approval processes typically depend on predefined conditions, approval rules, and workflow configurations.

Intelligent Approvals introduces generative AI capabilities to interpret business approval policies and help automate approval processes.

One important capability is translating uploaded approval policy documents into AI-generated approval rules.

Potential applications include