ServiceNow Knowledge26 Day 2 Keynote: From Patchwork Enterprise to Agentic Business

If your enterprise has already deployed AI agents and you’re still not seeing results, Day 2 of ServiceNow Knowledge 2026 was built for you. 

The keynote started by addressing the problem that most organizations don’t recognize – declining enterprise AI maturity, which has dropped 20% year over year in spite of increasing investments towards AI. Amit Zavery, in his keynote, highlighted that only 19% of businesses are getting real results from AI. This is because the foundation on which enterprises deploy AI agents is disjointed. 

Enterprises today work with a set of disconnected systems and applications, each having its own security rules, its own data model, and workflow logic that cannot communicate with each other. The gaps between the foundation and AI deployment are unveiled on day 2 of the ServiceNow Knowledge announcements. Day 2 revealed that the entire ServiceNow architecture is organised into four layers: Sense, Decide, Act, and Secure. Together, these layers are called the “blueprint for agentic business”. 

This blog uncovers all the major announcements across these four layers, and what each one means for how your enterprise runs AI.

ServiceNow Knowledge 2026 Day 2 Keynote Highlights

Day 1 keynote of ServiceNow Knowledge focused on the AI chaos. Day 2 opened with the answers that allow enterprises to look more deeply into their agentic infrastructure and find gaps in how they deploy and govern AI. 

ServiceNow Knowledge announcements gave a complete architectural insight into why enterprise AI keeps failing at scale. ServiceNow divided its architecture into a four-layer platform, including Sense, Decide, Act, and Secure. 

On day 2, the keynote covered a lot of announcements that highlight what an agent AI roadmap of an enterprise should look like. Here’s the breakdown –

• Workflow Data Fabric expanded to 250+ zero-copy connectors by adding 100 new connectors. With this, ServiceNow doesn’t copy or move the data but rather brings live queries directly to your data warehouse, data lake, or transactional systems.
• Data Catalog powers Autonomous Data Governance in ServiceNow, ensuring data Quality, Observability, Enrichment, Harmonization, and Policy. It gives enterprises the governed data foundation agents need to operate reliably.
• The major announcement unfolded in RaptorDB Pro, which highlights an upgrade to the core database layer of ServiceNow that runs transactional and analytical queries on the same live dataset at the same time.
• ServiceNow announced 20 new pre-built, role-scoped agents in a marketplace inside its platform. Enterprises can calculate the projected ROI of their agent against their operational data. 
• The platform revealed its Enhanced Security Foundation, where three inputs feed the AI Control Tower: Assets, Access, and Knowledge.   
• The biggest announcement of all is one year of free access to ServiceNow’s AI Control Tower, where every model, agent, and workflow will be visible, governed, and auditable through a unified interface.
• ServiceNow processes over 80 billion workflows each year, making it impossible to conduct manual security reviews. The partnership of ServiceNow with AWS Security brings the infrastructure and threat intelligence layer that makes automated governance at that scale highly reliable.

SENSE: Connecting, Governing, and Activating Enterprise Data

The keynote speech on day 2 of ServiceNow Knowledge highlighted that many enterprises do not check the platform or data readiness and go straight to deploying agents. It is exactly where they fail to scale their agentic enterprise. The disconnected enterprise data across hundreds of systems, when left ungoverned, becomes unreliable for AI to act on. 

ServiceNow fixes this with Workflow Data Fabric, Autonomous Data Governance, Autonomous Data Analytics, Context Engine, and RaptorDB Pro. 

Let’s break down each one of them:

a. Workflow Data Fabric

Workflow Data Fabric is how ServiceNow connects that enterprise data to external systems and agents without creating another copy of it. In ServiceNow’s announcements, there are 100 new Zero-Copy Connectors, bringing the total to 250+. They cover data warehouses, data lakes, and transactional systems across any cloud or on-premise environment. They query the same records your system holds, without manipulating or repeating the data.  

b. Autonomous Data Governance

ServiceNow announcements we uncovered on day 2 include Autonomous Data Governance. AI agents cannot trust systems and applications that have connected data without governance in place. Autonomous Data Governance covers 5 capabilities:

• Quality handles automated data profiling, scoring, and cleansing.
• Real-time observability, pipeline monitoring, and anomaly detection.
• Enrichment helps AI agents complete their context, which is possible by connecting them with external and internal datasets.
• Harmonization creates a single source of truth across agents and workflows.
• Policy handles the definition, management, and enforcement of data governance rules at the platform level. 

c. Autonomous Data Analytics and the Context Engine

Once enterprises connect and start governing data, the next question is whether your agents can actually understand it well enough to act on it without being asked.

Enterprises can convert governed data into intelligence using Autonomous Data Analytics. It combines Autonomous insights with Autonomous data apps and the Context Engine. Hence, with ServiceNow, you can close the gap between what data means and what the agent understands by keeping Context Engine in place.

d. RaptorDB Pro

RaptorDB Pro is the database layer underneath Workflow Data Fabric,  Autonomous Data Governance, and Autonomous Data Analytics. The upgrade in RaptorDB Pro announced at K26 addresses a specific bottleneck that matters a lot for agentic workloads. 

ServiceNow unveils two major capabilities in its database engine: Live Connect with SQL API, allowing external systems to send queries using standard SQL, and Live Archive, handling data archiving alongside the core engine, eliminating the need to pull the data out of the system entirely. 

DECIDE: Context Engine and the Partnership Layer

The keynote speech highlighted that AI needs access to data, but it should also have a relational understanding of people, assets, policies, and history to make responsible and reliable decisions. The Decide layer of the ServiceNow addresses the problem of ‘context blindness’. 

Why Context Engine Matters for ServiceNow?

Context blindness occurs when an AI agent does not have the context of the problem, and hence it cannot perform productively. The context engine in ServiceNow is a continuously updated operational model, connecting people, assets, roles, services, policies, and workflow histories into a single interface. 

Enterprises can use any model they are already using, like OpenAI, Anthropic, Google, Mistral AI, Nvidia, etc, to plug into the context engine and leverage the governance and security of your enterprise without requiring third-party tools or any configuration. Hence, the ServiceNow platform brings intelligence with built-in context, boundaries, and accountability.

Google Cloud and Gemini Enterprise partnership

The deepest integration shown on stage at K26 was the Google Cloud partnership, and it covered 3 things that matter for enterprises running complex multi-system environments.

The full-stack capabilities of this partnership include Google models, asset intelligence, and agentic capabilities that connect across every ServiceNow product to deliver first-party experiences at scale. This partnership enables a unified governance as Gemini Enterprise can be integrated directly with ServiceNow’s governance layer, giving a unified governance control to enterprises.  

ACT: Autonomous Workforce, Build Agent, and Live Demos

The Act layer in ServiceNow is where AI stops suggesting and begins executing actions. All the other layers feed into the Act layer, where the Sense layer ensures the data is live, properly governed, and reliable, while the Decide layer ensures that AI understands the context before it moves into execution.

Think of it like this – Most enterprise AI is operating as an advisory layer. They can search for insights and recommend next best actions, but a human agent has to intervene to actually get the work done. Whether it’s triggering the urgent approvals, opening the right system, or updating the records, the act layer closes the gap between suggestion and end-to-end execution. 

Autonomous Workforce: AI Specialists working alongside your teams

ServiceNow announced 20 new AI Specialists at K26, covering Finance, IT, Human Resources, Procurement, Security and Risk, CRM, and Legal. They’re available in a marketplace inside the platform described on stage as your marketplace for all things agentic. 

ServiceNow allows enterprises to operate inside the Act layer through Autonomous Workforce, which is the AI specialists that can handle end-to-end processes alongside human agents. The AI specialists in the Act layer are role-scoped, which means they will operate inside the same permission boundaries as the human worker performing in that role. 

These AI specialists cannot change access; they cannot self-escalate, and every action taken by them requires the same ACL framework that governs other employees in your enterprise. These AI specialists can handle routine work like compliance checks, case routing, benefits enrolment, access provisioning, and much more.

SECURE: AI Control Tower as the Governance Layer for the Agentic Enterprise

The ServiceNow keynote on day 2 highlighted the importance of creating secure boundaries and guardrails for AI agents to operate in. 

Enhanced Security Foundation: Assets, Access, and Knowledge

ServiceNow announced the Enhanced Security Foundation, which extends governance across every asset class. It works through 3 inputs that together give the Control Tower complete operational visibility.

• Assets extend governance beyond traditional IT into the full cyber-physical environment: IoT devices, operational technology, code repositories, cloud infrastructure, and AI agents themselves.
• Access feeds the Control Tower with every identity, role, group membership, and permission set across your enterprise. Human workers, machine identities, and AI agents all operate under the same ACL framework.
• The AWS Security partnership reinforces this further. In May 2026, ServiceNow and AWS integrated AWS Security Agent directly with ServiceNow’s Security Incident Response and Vulnerability Response modules, enabling autonomous threat detection, on-demand penetration testing, and remediation.
• Knowledge is where policies, regulations, vulnerabilities, and known attack paths live. 

Bottom Line

If your enterprise data is still disconnected across systems, running AI agents on top of it will produce unreliable outputs, regardless of which AI model powers them. ServiceNow Day 2 made one thing clear: enterprise AI needs to access data in real time with a trusted governance framework underneath it. The real question to ask your team is whether your data is connected and governed in real time, or not.

Most enterprises that aren’t seeing results from AI deployments have a Sense layer problem, not a model problem. Cyntexa’s ServiceNow specialists work with enterprises to assess AI readiness and identify the right entry points into the Sense, Decide, Act, Secure architecture. 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.