ServiceNow Knowledge26 Day 3 Keynote: When the Platform Becomes the Governance

If you followed Day 1 and Day 2 of Knowledge 2026, you already know the problem ServiceNow has been building toward solving – Governing the AI Chaos.

Knowledge Day 3 keynote began with the announcement of the ServiceNow Australia release, which is their answer to all things AI. Enterprises deploying dozens of AI agents across their systems with no governance layer are exposing themselves to a multitude of risks. What they need is central visibility, accountability, and a reliable way to measure what they’re delivering. 

This blog covers every major announcement from Knowledge Day 3, what each capability actually does, and what it means for enterprises and developers trying to move from AI deployment to AI that actually holds up. Let’s get started.

What Was Announced At ServiceNow Knowledge Day 3

The major announcements came from the Platform Keynote delivered by Heath Ramsey (Group VP AI Platform, ServiceNow), the Developer Creator Keynote, and the Data and Analytics session –  

• The Platform Keynote announced the Australia release, introducing AI Agent Advisor, AI Agent Evaluator, Knowledge Center, Dynamic Guidance, Intelligent Approvals, and AI Control Tower expansions.
• The Developer CreatorCon Keynote, which introduced Build Agent, external tool governance using the ServiceNow skill plugin, WarpSpeed, and EmployeeWorks.
• The Data and Analytics session highlighted how ServiceNow builds with context, 350+ zero-copy connectors, full MCP support, AI-native Data Fabric, RaptorDB Pro, Autonomous Data Governance, and Partner Passport. 

We will look into each of these announcements in detail and understand what it means for enterprises scaling AI. 

The Australia Release: What It Is and Why It Matters Now

The keynote on Knowledge Day 3 announced the newest release of the ServiceNow platform, i.e., Australia. It is one of the most significant releases so far because it’s the first release where AI governance is integrated within the architecture, not as an add-on. 

Australia release is powered by 4 superpowers ServiceNow announced at Knowledge: Sense, Decide, Act, Secure.

SENSE: Find What to Automate, Then Deploy It

AI Agent Advisor

ServiceNow announced AI Agent Advisor in their Australia release, through which enterprises can find automation opportunities for the workflows that keep lagging inside their systems, waiting for approvals or human support. 

Such processes can be executed faster. To resolve this, AI Agent Advisor proactively determines the broken workflows and suggests actions that will assist in fixing delays, like escalating requests with AI agents, automating hand-offs, etc.

AI Agent Evaluator

AI Agent Evaluator is the newest addition in the ServiceNow Australia release. It tests agents against different conditions and parameters and indicates completion rates, decision quality, and accuracy, which can be analyzed by enterprises before going into production. 

If enterprises are putting this kind of baseline evaluation in place, they cannot succeed at large-scale AI deployment. Agent Evaluator eliminates interpretation. It means, instead of wondering why your agent performed a certain way, you take complete control and test its capabilities against multiple use cases using ServiceNow’s AI Agent Evaluator.  

AI Desktop Actions

AI Desktop Actions allows the ServiceNow platform to dynamically create AI agents that handle repetitive, manual tasks on a user’s desktop. It eliminates the need to perform manual tasks like copy-pasting, screen navigation, repetitive data entry between systems, etc. 

DECIDE: Giving AI the Right Information at the Right Moment

Knowledge Center

In the Australia release, ServiceNow users have a Knowledge Center. It is a dedicated workspace for knowledge managers to find duplicate articles, identify content gaps, and merge overlapping documentation, all using AI. 

The Knowledge Center is a part of the broader Decide layer of ServiceNow, which treats the quality of information as the foundation for the quality of AI decisions. It simply means that AI decisions are as good as the data it is working on. Hence, enterprises must focus on getting the Knowledge Center in place before they scale AI agents.    

Dynamic Guidance

ServiceNow announced Dynamic Guidance, which is a voice-activated AI assistant built directly into the platform. It helps users understand what they’re looking at and what to do next, in real time. 

Think of it like this: when a new user accesses ServiceNow, they don’t understand what they should work on, or even how to explore instances. 

The presenter performed a live demo on stage and asked the Dynamic Guidance for help. The assistant walked him through key metrics, guided him to open the menu, personalise the view, and reorganize the data table, all using natural language commands. New users can navigate ServiceNow without leaving the platform, opening a separate tool, or waiting for training. 

ACT: Automation That Works on Underlying Policies

Intelligent Approvals

Enterprises must think about how approval processes work inside their teams. They must act on designing policies that will govern the approval and rejection within the system. 

With Intelligent Approvals introduced at Knowledge Day 3 Keynote, enterprises can feed actual policy documents into the ServiceNow platform, which becomes the single source of truth for agents. 

Heath Ramsey demonstrated this live with two travel requests. For instance, A trip to Las Vegas for a company event is automatically approved by the system because the policy covers it. On the other hand, a request for a client dinner in Berlin is automatically rejected, because the policy doesn’t cover international client entertainment. 

SECURE: AI Control Tower Gets a Governance Upgrade

One of the major announcements was made regarding the governance upgrade in AI Control Tower. 

In ServiceNow, AI Control Tower gives you a single place to see everything that’s happening across your entire AI environment, not just inside ServiceNow. Enterprises can track the full inventory of AI assets, review regulatory risk tracking, assess lifecycle status, and see how much business value AI agents are generating. All of this data is now available inside one view – AI Control Tower.

ServiceNow Build Agent: The Biggest Developer Announcement of Knowledge 2026

What Build Agent Does

ServiceNow Build Agent allows users to create ready-to-deploy applications from scratch using vibe coding. It is an AI-powered development assistant built directly into the ServiceNow platform.

Build Agent can deploy full-stack applications by understanding context from conversational language, reading code scripts, proposing architectural improvements, finding bugs, tracing external integrations, and building new features as enterprises scale. 

What it Means for Developers and Enterprises

Developers don’t have to leave the AI agents they currently work with. They can use the ServiceNow skill plugin to bring ServiceNow platform knowledge directly into external development tools like Cursor and Claude Code. 

For instance, an agent in Claude Code knows how to compile a sub-flow, how to wire inputs and outputs, and how to use the fluent API, because that knowledge comes from ServiceNow itself. The platform’s governance travels with code and applies across all AI agents your teams work with.

WarpSpeed: A Governed Build Environment for Non-Developers

As announced in the Developer CreatorCon Keynote, WarpSpeed is a governed hosting environment where non-developer users can build applications that run on real production data. It inherits permissions from the underlying governance layer. 

WarpSpeed helps clear out developer backlog by allowing even non-developers to take charge and build applications using conversational prompts and instructions. In addition, it offers governed hosting, app deployment review, and a path to deployment using a controlled and governed process.  WarpSpeed allows non-developers to request and use apps even without a ServiceNow developer license. 

EmployeeWorks: Connecting Everything Built to Everyone 

Developers build apps anywhere, govern them through ServiceNow, and reach their entire organization through EmployeeWorks as the unified delivery surface. Hence, it acts as a layer connecting everything built on the ServiceNow platform to the employees who need to use it.

EmployeeWorks creates a clear delivery path for employees by allowing them access to apps built in WarpSpeed, agents deployed through AI Agent Studio, and workflows governed through AI Control Tower. 

What’s New in Data and Analytics

Zero-Copy Connectors 

The session on Data & Analytics raised one question for its audience – where does the data actually come from?

AI agents are only as good as the data they can reach and trust. And as highlighted from the session, ServiceNow found gaps in the enterprise data connectivity. Organizations continue to deploy multiple AI agents across distributed systems and environments. With this, they create a serious discontinuity between their systems, applications, and AI agents. When agents don’t have a full context of the problem, they generate ambiguous outputs.   

Hence, ServiceNow announced more than 350 zero-copy connectors, allowing agents to work with data wherever it lives without requiring its migration.

MCP Expansion 

MCP (Model Context Protocol) is a standardized protocol that allows AI agents to interface with external data sources without requiring a custom integration built for each one. 

ServiceNow announced that they are expanding MCP support. It means that as new data sources come online, they can be connected using the same protocol rather than requiring custom integration work each time. 

Most enterprises have data scattered across dozens of systems, but with zero-copy connectors and MCP support, they don’t need to move their data. AI agents will work where their data is while under ServiceNow governance and platform control.  

Highlights From The MVP Panel at Knowledge Day 3

The MVP panel in the Developer Creator session at Knowledge 2026 included one of the most honest conversations of the day. The focus of this session remained on –

• Enterprises mustn’t treat governance as an afterthought. It isn’t a feature you can add after App/AI agent deployment. 
• In addition to fixing bugs, Build Agent captures the reasoning behind every decision on the platform. It performs with context and reasoning, which is why the next developer who works on the code doesn’t spend months breaking and building it.
• Enterprises can only trust AI when they trust the underlying data. So, to strengthen their AI and scale responsibly, they need a robust data foundation. ServiceNow’s AI Control Tower is designed to provide that trust visibility into the system. It ensures that what the agent sees is accurate, what it does is within defined boundaries, and every decision it makes is visible and auditable in one place.

Bottom Line

ServiceNow Australia release is not about adding more AI features to the platform. It’s about governing the ones you already have and making them trustworthy enough to act on.

If you’re looking to assess your current ServiceNow environment against the architecture shown this week, Cyntexa’s ServiceNow specialists can help you identify the right entry points into the Sense, Decide, Act, Secure framework. 

Connect with our team to get started.

AUTHOR

Vishwajeet Srivastava

Salesforce Data Cloud, AI Products, ServiceNow, Product Engineering

Co-founder and CTO at Cyntexa also known as “VJ”. With 10+ years of experience and 22+ Salesforce certifications, he’s a seasoned expert in Salesforce Data Cloud & AI Products, Product Engineering, AWS, Google Cloud Platform, ServiceNow, and Managed Services. Known for blending strategic thinking with hands-on expertise, VJ is passionate about building scalable solutions that drive innovation, operational efficiency, and enterprise-wide transformation.