Combined AI Control Tower release notes for upgrades from Xanadu to Australia

Consolidated page of all release notes for AI Control Tower from Xanadu to Australia.

How to use this page

To help you prepare for your upgrade, we have combined the cross-family AI Control Tower release notes onto one page. Read this summary of the new features, changes, and updated information for your product from Xanadu to Australia.

Tip:

If there were no updates for a release notes section in a certain family release, we included a short note for your reference. For example, if a product did not have any updates in Tokyo, the row says "No updates for this release."

Important information for upgrading AI Control Tower to Australia

Before you upgrade to Australia, review these pre- and post-upgrade tasks and complete the tasks as needed.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	General availability release, no upgrade.
Zurich	No updates for this release.
Australia	Not applicable.

New features

Between your current release family and Australia, new features were introduced for AI Control Tower.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	Yokohama Patch 6 Health tab in AI Control Tower Monitor and evaluate the effectiveness of offensive content and prompt injection guardrails active on your AI assets. Evaluation tab Measure and improve the quality of interactions with virtual agents using the Evaluation tab. AI model providers Explore AI model providers Enable choice for third party model providers powering ServiceNow® skills and agents. Yokohama Patch 3 AI Governance A single pane view of the AI inventory, its state, and its risk and compliance posture. Lifecycle to manage AI asset onboarding and deployment. Helps user oversee and manage AI Asset inventory's risk profile with regard to enterprise policies and global regulations, as defined by the user, with a focus on privacy, data governance, and ethical AI. AI Case management to oversee AI asset-related inquiries and cases, enabling faster response and improved tracking. Multi-instance management to synchronize AI asset inventory from sub-prod to prod instances to initiate governance early in the build process. Control settings to block only ''other'' skills in Now Assist AI deployment pending approvals. AI Governance AI Steward role- Facilitate and coordinate governance activities between innovation, legal, security, risk and compliance teams. AI Asset inventory- Unified data model on the ServiceNow AI Platform to catalog AI Model, datasets, prompts, and other related artifacts including Now Assist and AI leveraging Generative AI Controller. AI skills Approvals- Review and approval flows for Now Assist skills and other related assets like AI Models and AI datasets deployed through Now Assist or generative AI Controller. AI Control Tower Workspace- Intuitive workspace to surface governance tasks, reports, inventory, and insights.
Zurich	Zurich Patch 7 Security and privacy tab Measure whether your model's output or behavior potentially violates predefined LLM guardrail policies using the Data integrity incident detection chart. Review potential threats in AI agent output in Agent goal deviation, Output with PII detected, and Agentic output injection detection charts. Monitor MCP server access by AI Gateway with these new charts: Clients connecting to MCP servers, authorized access attempts, and failed access attempts. Data section on Configurations page Enable and set up data integrity incident detection, agent goal deviation, and output screening metrics to measure the integrity of your data model and potential threats in LLM output. Manage agentic AI system life cycles Create AI system assets to track and manage the complete life cycles of your agentic AI systems, from onboarding to deployment. Gain comprehensive insight into each agentic AI system and take any necessary actions to successfully complete each life-cycle stage. By managing the life cycles of your agentic AI systems, you can extend their lifespans, reduce downtime, and optimize licensing costs. Define the use and purpose of an AI system Specify the intended use and purpose of an AI system. Provide insight into who is using the AI system, what the AI system is being used for, and how the AI system works and provides value. This information can help you determine the benefits and risks that are associated with the AI system. For more information on classifying AI systems based on regulatory risk at intake by applying a configured Risk Assessment Methodology (RAM), see Assessment templates and risk assessment methodologies and Request an AI use case. Associate additional related AI asset types with AI systems Associate the following additional related AI asset types with your AI systems: If an AI system has an Asset type of Generative AI or Agentic AI, you can associate it with any of its supported components or subsystems. If an AI system has an Asset type of Agentic AI, you can associate it with any of its integrated AI tools. Create change and offboarding requests for additional AI asset types Create change requests for the following additional AI asset types: AI systems with an Asset type of Agentic AI Datasets In addition, create offboarding requests for the following additional AI asset types: AI systems with an Asset type of Agentic AI AI models Datasets MCP servers Zurich Patch 4 Security and privacy tab Identify ServiceNow® AI assets that impact your security posture using the ServiceNow® AI security score and AI insights. AI insights highlight key metric changes, recommend next steps, enabling you to quickly understand the impacts and take action. Access and monitor security for AWS Bedrock agents running as privileged users, autonomous vs. supervised tools, and dormant agents. Monitor sensitive data detection, prompt injection, and offensive content metrics to help identify and mitigate AI-driven security and compliance risks before they impact workflows or expose sensitive information. AI security tasks are created automatically from Dormant AI systems metrics to streamline your workflow and quickly resolve issues. You can also create AI security tasks directly from more areas, such as access issues and privileged AI agents metrics. Review Autonomous vs. supervised systems metrics based on AI tools. Previously, the metrics were based on workflows. Show Access issues metrics for only those agents with issues. Previously, agents with issues and no issues were shown. See more details about agent access issues in the access map to help you troubleshoot quickly. For example, you can see the user ID of the user who executed the agent and the workflow and tool associated with the access issue, if applicable. AI Gateway tab The AI Gateway tab shows the metrics at the MCP server level, listing all connected MCP servers along with the total number of transactions for each server and its success rate. Data section on Configurations page See a read-only view of your data privacy configuration for sensitive data patterns in the Data privacy page. Use this page as a quick reference when troubleshooting sensitive data charts. AI Task section on AI assets page Review all AI security tasks for your instance in the All Security Tasks page. You can also create an AI task on this page. Enhance control of AI asset life cycle through change and offboarding requests Enhance the management of deployed AI assets by using the change request workflow to make necessary edits to AI assets that have already undergone review and onboarding. Furthermore, facilitate the retirement of AI assets by submitting an offboarding request, ensuring a structured and controlled process for removing assets that are no longer needed or have been superseded. Zurich Patch 1 Health tab in AI Control Tower Monitor and evaluate the effectiveness of offensive content and prompt injection guardrails active on your AI assets. Zurich Early Availability AI strategy with Strategic Planning(Requires SPM) Monitor and track your AI strategies and associated goals and targets. Track the costs of your AI projects, epics, and demands. Monitor key project risks, issues, decisions, actions, and changes. AI strategy with Goal Framework (Requires SPM) Monitor and track your AI strategies and associated goals and targets. AI connections Discover and add AI agents and related models and tools to the AI inventory through integration with AWS Bedrock. Discover and add AI agents and related models and tools to the AI inventory through integration with Azure AI Foundry. Configure AI discovery setup and visibility of connections. Value tab Gain insights into the value realized from AI skills and features. The Value insights dashboard page gives you insights into the estimated productivity gains as a result of using AI systems. Define and measure value relevant to your AI systems using customizable value templates. Perform calculations and approximations for the read and write time saved by users using AI systems by using data points and timestamps from the invoking records. Understand the key usage and performance indicators that help you evaluate the adoption of Now Assist in your organization. Provide insights on success rate visualization by department, country, and AI assets along with the indicators for task closure efficiency. Value templates Create, manage, and use templates from a Global Template Repository across multiple AI assets, so that you can standardize and streamline your AI experience, and track usage of AI systems, use cases, and skills more effectively. Enable users to edit, view, and create customized value templates by enabling a value template assignment experience in inventory records. Provide transparency to value calculations of each AI system through value templates. Perform calculations and approximations for the read and write time saved by users using AI systems by using data point and timestamps from the invoking records. Risk and compliance tab Display the risk classification of AI assets and the compliance posture for selected authority documents and policies through the Risk and compliance tab. This tab provides visibility into AI systems, models, and datasets. The Risk overview section uses visual charts to categorize AI assets. Review adherence to frameworks in the Compliance overview section. For example, adherence to the NIST AI Risk Management Framework or the EU Artificial Intelligence Act, presenting scores based on citations and control attestations that are set by the customer. You can filter data by authority documents or policies, view overall compliance percentages, and identify critical issues and AI cases tied to items deemed non-compliant by the customer. Build trust in the AI asset inventory to effectively manage AI-related risks across the enterprise. Drive enterprise-wide risk visibility by aggregating individual scores into a consolidated AI risk profile to support informed mitigation decisions. Display real-time residual risk scores on the home page to help practitioners identify high-risk AI assets and prioritize mitigation actions. Use enhanced impact assessment templates to help manage and oversee compliance with regulatory requirements. Perform bulk control attestations using Core UI to validate multiple controls across AI assets, improving efficiency for large-scale assessments. Adopt a proactive approach by leveraging comprehensive AI risk and compliance scoring across the entire AI asset inventory. AI cases tab Gain a centralized overview of all your AI asset cases and inquiries by using the AI cases. On the AI cases tab, you see a list of records that include the case details such as the status, priority, owner, and timeline of your AI cases. You can monitor the progression of a case, stay informed about ongoing investigations, follow up on pending actions, and help to ensure timely resolutions. On this tab, you can also find filtering and sorting options that help you to prioritize cases that require immediate attention. Use the enhanced home page to access a single, unified view of all AI-related cases and inquiries. Track, manage, and respond to AI-related cases more efficiently through centralized case visibility. Security and privacy tab Review your AI security health metrics in the Security and Privacy tab. Use the access map for a comprehensive overview of agentic workflows, tools, and agent details. The map helps to show how AI agents interface with workflows and tools to accomplish a task. Additionally, you can analyze current AI access, investigate ongoing access issues, and review your AI usage metrics. Control which third-party models OEM by ServiceNow are enabled for Now Assist AI implementation and how they’re used. Specify additional details during AI asset creation Use the following fields to specify additional details about your AI systems, AI models, prompts, and datasets when you create AI assets: AI systems: Managed by License details Supported locations AI models: Managed by License details Supported locations Prompts: Managed by Datasets: Managed by Creation type Department Dataset creation date
Australia	Australia Patch 2 Publish ServiceNow agents to Microsoft Agent 365 Publish the ServiceNow Agents to Microsoft Agent 365 ensuring the ServiceNow agents are sent to external registries. Service Graph Connectors for AI Control Tower AI Service Graph Connector for Databricks discover AI agents and import to AI Control Tower from Databricks environment. Australia Patch 1 Security & privacy tab in AI Governance Customize the AI asset security score by weighting LLM guardrail categories that comprise the score. The score formula was changed to an average across all AI assets. The score was renamed to the AI asset security score. Measure whether your model's output or behavior potentially violates predefined LLM guardrail policies using the Data integrity incident detection chart. Review potential threats in AI agent output in Agent goal deviation, output with PII detected, and Agentic output injection detection charts. Monitor MCP server access by AI Gateway with these new charts: Clients connecting to MCP servers, authorized access attempts, and failed access attempts. The Prompt injection, Offensive content, and Sensitive data tabs have been removed and replaced by the Access and Guardrails tabs. Metrics have been reorganized into those two tabs. In Configurations, under Data, the Data privacy tab was renamed to Security & privacy. In that tab, the data leak detection and anonymization section was renamed to sensitive data input and anonymization. Data section on Configurations page Enable and set up data integrity incident detection, agent goal deviation, and output screening metrics. These metrics measure the integrity of your data model and potential threats in LLM output. Manage agentic AI system life cycles Create AI system assets to track and manage the complete life cycles of your agentic AI systems, from onboarding to deployment. Gain comprehensive insight into each agentic AI system and take any necessary actions to successfully complete each life-cycle stage. By managing the life cycles of your agentic AI systems, you can extend their lifespans, reduce downtime, and optimize licensing costs. Define the use and purpose of an AI system Specify the intended use and purpose of an AI system. Provide insight into who is using the AI system, what the AI system is being used for, and how the AI system works and provides value. This information can help you determine the benefits and risks that are associated with the AI system. For more information on classifying AI systems based on regulatory risk at intake by applying a configured Risk Assessment Methodology (RAM), see, AI Risk and Compliance release notes Assessment templatesand Risk assessment methodologies. Associate additional related AI asset types with AI systems Associate the following additional related AI asset types with your AI systems: If an AI system has an Asset type of Generative AI or Agentic AI, you can associate it with any of its supported components or subsystems. If an AI system has an Asset type of Agentic AI, you can associate it with any of its integrated AI tools. Create change and offboarding requests for additional AI asset types Create change requests for the following additional AI asset types: AI systems with an Asset type of Agentic AI Datasets In addition, create offboarding requests for the following additional AI asset types: AI systems with an Asset type of Agentic AI AI models Datasets MCP servers ServiceNow product tiers The ServiceNow AI Platform now brings you a new AI experience with three licensing tiers available: Foundation: AI basics to deliver insights Advanced: AI to boost productivity across relevant use cases Prime: Act autonomously with all AI assets and create your own Depending on your entitlements, you will have access to certain application features, generative AI skills, agentic workflows, and AI agents. Early availability AI connections AI connections are created using AI Service Graph Connectors. AI connections are a combination of hyperscalers, AI apps, and agentic AI frameworks. The following AI Service Graph Connectors are available from March 2026 AWS Microsoft- Azure Foundry and Copilot Google Cloud Platform (GCP) Vertex AI n8n LangGraph Salesforce Managed and unmanaged AI assets Managed assets benefit from AI Control Tower features such as governance, lifecycle management, value assessment, risk classification, security, and privacy. Unmanaged assets, on the other hand, do not have access to these AI Control Tower capabilities. AI Gateway AI Gateway offers MCP Global Clients, which can be used across all servers. A Gateway offers MCP Catalog to choose while adding MCP servers. MCP server can be added to an AI Asset inventory from AI Control Tower.

Changes

Between your current release family and Australia, some changes were made to existing AI Control Tower features.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	Yokohama Patch 11 Changes to Now Assist usage measurement Starting with Yokohama Patch 5, Now Assist usage measurement is transitioning from a 365-day look-back model to a 365-day burn-down model, with usage resetting at the contract anniversary date. For more information, refer to KB KB2704710: Now Assist Usage - Overview & New Measurement Logic. Some Now Assist skills are turned on by default The new default behavior works as follows: New customers: When you install a Now Assist product, designated skills are turned on automatically. Existing customers who are upgrading (starting with Yokohama Patch 11): Any previously unconfigured skill is turned on automatically (the skill was never configured and turned on, then turned off again). Previously configured skills that were turned on, then off, remain inactive.
Zurich	Zurich patch 7 Security and privacy tab The Autonomous vs. supervised AI tools chart has been removed. The Prompt injection, Offensive content, and Sensitive data tabs have been removed and replaced by Access and Guardrails tabs. Metrics have been reorganized into those two tabs. In Configurations, under Data, the Data privacy tab was renamed to Security & privacy. In that tab, the data leak detection and anonymization section was renamed to sensitive data input and anonymization. Zurich Patch 5 Changes to Now Assist usage measurement Starting with Zurich Patch 5, Now Assist usage measurement is transitioning from a 365-day look-back model to a 365-day burn-down model, with usage resetting at the contract anniversary date. For more information, refer to KB KB2704710: Now Assist Usage - Overview & New Measurement Logic. Changes in Zurich Patch 4 The AI asset inventory plugin structure has been updated. Product owner view: Added a role called AI asset owner [sn_ai_asset_mgmt.ai_asset_owner], which enables the Product Owner view experience with a personalized home page and enhanced visibility into AI assets to simplify task management. AI discovery: The Innovation lab store application (AWS AI discovery plugin) is decommissioned. Uninstall the AWS AI discovery plugin prior to installing the AI discovery plugin (sn_ai_disc). AI cases management has moved under the AI cases tab on the AI Control Tower home page.
Australia	Australia Patch 3 Additional regulatory frameworks in the AI Risk and Compliance content pack After AI Risk and Compliance is updated to version 22.3.0 and the new frameworks are activated, authority documents, agency mappings, and citations for the Transparency in Frontier Artificial Intelligence Act (SB 53) and the Colorado Artificial Intelligence Act (SB 205) appear in the compliance posture and related views on the Risk and compliance tab. For more information, see AI Risk and Compliance release notes, Content pack, Activate or update the Transparency in Frontier Artificial Intelligence Act (SB 53), and Activate or update the Colorado Artificial Intelligence Act. Impact assessment field auto-population After upgrading to version 22.3.5, if you have the AI asset owner [sn_ai_asset_mgmt.ai_asset_owner] or AI risk and compliance business user [sn_grc_ai_gov.ai_risk_and_compliance_business_user] role, the screening question responses that capture the AI system's intended use and operational context from the Use and Purpose section of the AI use case request form are automatically populated in the corresponding Use and Purpose fields of a new impact assessment. This synchronization reduces manual entry and helps ensure that impact assessment responses are consistent with the information submitted at intake. For more information, see AI Risk and Compliance release notes and Intake requests.

Removed

Between your current release family and Australia, some AI Control Tower features or functionality were removed.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	No updates for this release.
Zurich	No updates for this release.
Australia	Australia Patch 1 The Autonomous vs. supervised AI tools chart has been removed from the Security & privacy tab. Early availability Adding legacy AI connections via Service Graph Connectors (SGC) is deprecated. In AI connections, under Legacy connections, the New button has been removed to block users from creating new connections using SGC.

Deprecations

Between your current release family and Australia, some AI Control Tower features or functionality were deprecated.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	No updates for this release.
Zurich	No updates for this release.
Australia	No updates for this release.

Activation information

Review information on how to activate AI Control Tower.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	The AI Control Tower application is installed as part of the generative AI Controller.
Zurich	Install AI Control Tower by requesting it from the ServiceNow Store. Visit the ServiceNow Store website to view all the available apps and for information about submitting requests to the store. For cumulative release notes information for all released apps, see the ServiceNow Store version history release notes.
Australia	Install AI Control Tower by requesting it from the ServiceNow Store.

Additional requirements

If any additional requirements were introduced or changed for AI Control Tower we have noted them here.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	No updates for this release.
Zurich	No updates for this release.
Australia	No updates for this release.

Browser requirements

If any specific browser requirements were introduced or changed for AI Control Tower we have noted them here.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	The AI Control Tower application supports all the browsers.
Zurich	No updates for this release.
Australia	The AI Control Tower application supports all browsers.

Accessibility information

Review details on accessibility information for AI Control Tower, such as specific requirements or compliance levels.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	The AI Control Tower application supports all the platform accessibility features.
Zurich	Dark theme The new Coral theme includes a dark theme option for web and mobile experiences. This option is commonly used to alleviate eye strain and improve readability.
Australia	The AI Control Tower application supports all platform accessibility features.

Localization information

If there are specific localization considerations for AI Control Tower we have noted them here.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	No updates for this release.
Zurich	No updates for this release.
Australia	The AI Control Tower application is localized.

Highlight information

If there are specific highlight considerations for AI Control Tower we have noted them here.

Release	Release notes
Xanadu	No updates for this release.
Yokohama	Yokohama Patch 11 Review changes to Now Assist usage measurement. Some Now Assist skills, agents, and agentic workflows are on by default. Additional role configuration is required for agentic workflows and AI agents included with Now Assist applications. AI connections are introduced in AI Control Tower using Service Graph Connectors. AI connections are combination of hyperscalars, AI apps, and agentic AI frameworks. The AI Service Graph Connectors available from March 2026: AWS Microsoft- Azure Foundry and Copilot n8n GCP Vertex AI LangGraph Salesforce Yokohama Patch 6 Monitor the performance of guardrails enabled through Now Assist Guardian using the Health tab. Measure and improve the quality of interactions with virtual agents using the Evaluation tab. Yokohama Patch 3 AI Control Tower helps customers manage and oversee performance, risk profile & workforce transformation while also helping to seamlessly embed AI into enterprise strategy. Create an AI steward role. Use the AI Asset inventory to catalog AI-related artifacts. Use the AI skills Approvals to review and approval flows. Create a AI Control Tower Workspace.
Zurich	Zurich Patch 8Configure and create automation rules to set AI assets as managed assets. Zurich Patch 7 Use new security metrics to monitor your LLM and AI agent output for potential security and content policy violations, potential PII, and other potential threats. Gain visibility into MCP client-server interactions routed through this instance’s AI Gateway. AI assets—Including AI models, AI systems, prompts, datasets, and MCP servers can be categorized as either managed or unmanaged. Managed assets benefit from AI Control Tower features such as governance, lifecycle management, value assessment, risk classification, security, and privacy. Unmanaged assets, on the other hand, don’t have access to these AI Control Tower capabilities. AI connections are introduced in AI Control Tower using Service Graph Connectors. AI connections are a combination of hyperscalars, AI apps, and agentic AI frameworks. The AI Service Graph Connectors available from March 2026: AWS Microsoft- Azure Foundry and Copilot Google Cloud Platform (GCP) Vertex AI n8n LangGraph Salesforce Manage the end-to-end life cycles of your agentic AI systems. Define the intended use and purpose of an AI system so that you can determine its benefits and risks. AI Gateway offers MCP Global Clients, which can be used across all servers. AI Gateway offers MCP Catalog to choose while adding MCP servers. MCP server can be added to an AI Asset inventory from AI Control Tower. Zurich Patch 5 Review changes to Now Assist usage measurement. Zurich Patch 4 Identify ServiceNow® AI assets that impact your security posture using the ServiceNow® AI security score and AI insights. Access and monitor security for AWS Bedrock agents running as privileged users, autonomous vs. Supervised tools, and dormant agents. Monitor sensitive data detection, prompt injection, and offensive content metrics to help identify and mitigate AI-driven security and compliance risks before they impact workflows or expose sensitive information. See more details in the access map about agent access issues to help you troubleshoot quickly. Audit logs capture configuration changes made on Data, Approvals, and AI model providers categories. Discover AI assets built and deployed in Google Cloud Platform (GCP) Vertex AI, Copilot Studio, and Azure AI Foundry. AI Gateway enables enterprises to actively manage, govern, and observe their MCP traffic, ensuring secure operation of agentic workflows across enterprise boundaries. Zurich Patch 1 Monitor the performance of guardrails enabled through Now Assist Guardian using the Health tab. Measure and improve the quality of interactions with virtual agents using the Evaluation tab. Display data based on the chosen allowed model providers and the status of the fallback in the Impact Summary table on the AI model providers section. Synchronize AI agents automatically when an AI asset is synchronized. Enhance the Product Owner experience with a personalized home page, value management tools to manage AI investments, and enhanced visibility into AI assets to simplify task management. Evaluate AI productivity and adoption across the enterprise using defined value metrics and performance indicators to drive data-informed decisions and maximize AI impact. Access and security monitoring for ServiceNow® AI agents, especially around access issues, agents running as privileged users and dormant agents. Discover AI assets built and deployed in AWS Bedrock and Azure Foundry. Enable choice for third-party model providers powering ServiceNow® skills and agents. Access to aggregated risk scores to improve decision-making, manage risks, and help to promote ethical and transparent AI practices. Monitor performance, track progress, and make informed decisions related to your AI strategies, goals, targets, and the associated work from the AI strategy tab. Track costs of your AI projects, epics, demands, and track key project risks, issues, decisions, actions, and changes from the AI strategy tab. See AI Governance for more information.
Australia	Australia Patch 1 Customize the AI asset security score calculation to reflect your security requirements. Use new security metrics to monitor your LLM and AI agent output for potential security and content policy violations, potential PII, and other potential threats. Gain visibility into MCP client-server interactions routed through this instance's AI Gateway. Configure and create automation rules to set AI assets as managed assets. Manage the end-to-end life cycles of your agentic AI systems. Define the intended use and purpose of an AI system so that you can determine its benefits and risks. Early availability AI assets—including AI models, AI systems, prompts, datasets, and MCP servers can be categorized as either managed or unmanaged. AI connections are introduced in AI Control Tower using Service Graph Connectors (SGC). The AI model providers supported by ServiceNow contains providers such as Now LLM Service, AWS Claude, Now LLM LTS model, and so on. The AI model providers configured by your organization contains providers such as Perplexity, IBM Watson, and so on. AI Gateway offers Global MCP clients, which once created can be used across all MCP servers. A Gateway offers MCP Catalog to choose while adding MCP servers into AI Control Tower. For more information on AI Control Tower, see AI Governance.