By Alex Pozin and Samer Baroudi
Maintaining visibility and control becomes increasingly complex as organizations deploy more agents across multiple Microsoft 365 tenants, through a heterogenous mix of agent platforms, and across different teams or areas in the organization. Microsoft Agent 365 has introduced new capabilities that help IT, security, and business leaders govern this growing agent ecosystem with greater confidence. The July 2026 updates bring partner risk signals into a unified security view, improve oversight of usage-based AI spend, deliver organization-wide adoption insights, extend agent discovery and governance across ecosystem platforms, and enable centralized management across multiple tenants. Together, these innovations advance Agent 365 as the enterprise control plane for observing, governing, and securing agents at scale.
Expand agent risk visibility with signals from Cyera, Darktrace, Zenity
As agents take on more business-critical work, IT and security teams may need to move across multiple portals to understand risk signals spanning identity, data, and threat activity. This fragmented view can make it harder to see an agent's overall risk profile and determine where attention is needed. Microsoft Agent 365 addresses that challenge with a unified control plane that brings agent risk signals together, giving administrators a consistent place to assess and act on risk.
Microsoft Security delivers native, deeply integrated Agent 365-enabled risk insights and security value across identity, threat protection, data security, compliance, and device management. Today, we are extending visibility for customers who use third-party security solutions by supporting selected partner risk signals in the Agent Registry, beginning with Cyera, Darktrace, and Zenity.
"Microsoft is committed to delivering the industry’s most trusted and comprehensive AI governance and security experience as part of the Agent 365 promise. Through integrations with our security ecosystem partners, we enable customers to maximize the value of their existing security investments while gaining a unified, end-to-end view of risk signals across their agent ecosystem."
-- Nirav Shah, CVP, Agent 365 and Agent Cloud
Today, customers see agent risk signals within the Agent Details Security tab from across Microsoft Entra, Defender, and Purview. Now, when enabled by an administrator, Agent 365 can also ingest a partner-generated risk signal, matching it to the corresponding agent through Microsoft Entra Agent ID, and enriching your overall agent risk signal profile from within Agent 365. Microsoft Entra Agent ID serves as the identity anchor that allows supported partner signals to map cleanly to the right agent record. This keeps Agent 365 at the center of the governance workflow while giving customers a consolidated view of signals from the security tools they use.
Once an integration is active and an active risk signal is received for an eligible agent, these signals appear in the Agent Registry alongside Microsoft-generated risk signals.
For administrators, Agent 365 provides:
- One consolidated view - agent risk signals count brings together native signals from Microsoft Entra, Defender, and Purview with detections from supported partner solutions, helping administrators quickly identify agents that require attention.
- Risk prioritization – to help customers understand the overall risk picture for an agent, agent risk signals are prioritized into High, Medium, and Low categories. In addition, each risk signal type shows a number of occurrences, which can help identify repeat patterns requiring investigation or action.
- Consent by design - partner integrations are optional, and no partner data flows until an administrator enables partner risk signals under Agents > Settings in the Microsoft 365 admin center. Some partners may also require configuration in their own product.
This announcement reinforces Microsoft's integrated advantage and commitment to customer choice and meeting customers where they are. These optional partner integrations simply extend Agent 365 security risk visibility based on a customer’s current security solution investments.
"Cyera gives security teams the confidence to accelerate AI adoption by understanding exactly what each agent can reach, without putting sensitive data at risk. By bringing Cyera Agent Guardian's data risk signals into the Agent Registry, Microsoft Agent 365 gives administrators one consolidated view that combines native Microsoft signals with the data context our platform provides. Together, that's the model this next phase of enterprise AI adoption requires — governance that scales as fast as the agents do."
-- Tamar Bar-Ilan, Co-Founder & CTO, Cyera
“With Darktrace / SECURE AI, our behavioral approach builds an understanding unique to each organization of how AI agents operate and how their actions connect to people, data and infrastructure. As agents take on more business-critical work, that context is essential for uncovering subtle behavioral changes, unexpected interactions and emerging risks that static controls or known indicators may miss. Bringing those insights into Microsoft Agent 365 gives customers a clearer view of agent risk and the context to act with confidence.”
-- Ed Jennings, President and CEO, Darktrace.
“AI Agents are a reality in every enterprise today- what has started with AI agent pilots is now becoming full scale enterprise-wide deployments, and without a deep security pillar, it just can't happen. We are proud of this important first step in our partnership with Microsoft Agent 365- the Zenity integration with Agent 365 surfaces the critical security issues and real-time threats that AI agents are facing directly in Microsoft Agent 365. Enterprises can now use it to quickly prioritize and act on what matters most with regards to their agentic footprint."
-- Ben Kliger, CEO of Zenity
The Agent 365 partner API carrying these signals is available to security vendors across the ecosystem, and we anticipate the list of supported providers to expand in the future. As the agent ecosystem grows, Agent 365 will continue to bring the visibility, governance, and security organizations need to manage it with confidence.
Bring predictability to agentic AI spend with Cost Management
As organizations expand their use of agents, they need clear controls to understand consumption, manage budgets, and keep usage-based costs predictable. Cost Management in the Microsoft 365 admin center gives admins a centralized way to monitor and control Copilot Credit spend for agentic usage-based services. This release focuses on Cowork and Work IQ spend. New enterprise-ready updates improve the reliability of spending alerts, make policy assignment more consistent, strengthen the handling of per-user overages, and provide a clearer view of prepaid capacity across Microsoft 365 and Power Platform admin experiences. When a user belongs to multiple spending policies for the same service, the applicable policy is now selected using clear criteria: the highest per-user limit, followed by the largest overall policy limit and then the most recently created policy. The selected policy applies in full, helping admins anticipate which billing configuration and limits will govern each user without combining settings across policies. The updates also help protect shared capacity when a task exceeds a user’s individual limit while it is already running. With this release, Microsoft may allow that task to finish without drawing the resulting per-user overage from the policy’s shared pool or billing it, helping preserve credits for the rest of the group. Finally, admins gain a more accurate picture of capacity-pack credits available for use. Cost Management will distinguish credits already consumed or allocated through other admin centers, including capacity assigned to Power Platform environments or agents, so the amount presented during spending-policy setup more closely reflects the prepaid capacity that remains available.
Empower leaders and analysts with decision-grade org-wide agent insights
Now generally available, the Agent 365 Dashboard provides business leaders with a centralized view of agent adoption, engagement, and business impact across their organization. By consolidating telemetry from Agent 365 into a single reporting experience, the dashboard enables leaders to understand how agents are being used, identify where adoption is accelerating or stalling, and make more informed decisions about AI investments, deployment strategies, and change management efforts.
With built-in reporting and organizational filtering, analysts can evaluate usage across teams, business units, and geographies, while leaders can monitor key adoption metrics such as active users, retention trends, top-performing agents, and overall engagement patterns. Interactive drill-down capabilities provide deeper visibility into individual agent performance, helping organizations identify successful deployments, uncover optimization opportunities, and prioritize investments in the agents delivering the greatest business value.
Unify agent discovery and governance across agent platforms
Registry sync is now generally available in Microsoft Agent 365, giving organizations a centralized way to discover and inventory agents that exist beyond Microsoft platforms. As enterprises increasingly build and deploy agents across multiple agent development platforms, maintaining visibility becomes increasingly difficult. Registry sync addresses this challenge by allowing AI administrators to securely connect supported ecosystem agent platforms and synchronize agents and metadata into the Agent 365 registry on a daily or weekly cadence, establishing a single, authoritative system of record for their agent estate. With Registry sync, organizations can reduce blind spots ecosystem partner platforms by bringing externally developed agents into a unified discovery and management experience. Administrators can view synchronized agents alongside Microsoft-native agents, understand where agents originate, review key metadata, and gain a more complete picture of their growing agent landscape. Supported platforms at general availability include Amazon Bedrock, Google Cloud Gemini Enterprise Agent Platform, Anthropic Claude, Databricks Genie, Salesforce Agentforce. Additionally, Snowflake, Oracle OCI, and UIPath to follow soon.
Registry sync goes beyond visibility. Where supported by the underlying platform APIs, administrators can take agent-level governance actions directly from Agent 365, including lifecycle management actions such as deleting synchronized agents. This enables organizations to govern agents across platforms from a common management plane while continuing to use the development and runtime environments that best fit their business needs. By extending the Agent 365 registry beyond Microsoft boundaries, Registry sync strengthens Agent 365 as the enterprise control plane for agents. Whether agents are built by Microsoft, developed internally, or created on ecosystem platforms, organizations can now establish a more comprehensive view of agents operating across their organization, helping IT and security teams maintain visibility, reduce risk, and improve governance at scale.
Manage agents across multiple tenants from one place
Partners supporting multiple customers and enterprises operating complex multi-tenant environments both need a simpler way to maintain consistent visibility and control. With multi-tenant agent management, now in public preview, administrators can centrally view and manage agents across connected tenants from the Agent 365 in the Microsoft 365 admin center.
From All tenants > Agents, administrators can view a consolidated agent inventory, add custom agents, deploy agents across selected tenants, compare tenant-specific availability and permissions, review agent risk and activity, and block agents or change their availability where needed. They can also use the tenant switcher to enter a connected tenant through delegated access, without maintaining a separate administrator account in that tenant. Where applicable licensing requirements are met, administrators can also review agent risks. The experience builds on existing cross-tenant administration relationships and Granular Delegated Admin Privileges (GDAP). Partners configure these relationships in Partner Center, while enterprises can configure them through Microsoft Entra Tenant Governance. Available tenants and management actions are determined by the active tenant relationship and the administrator’s delegated role.
Multi-tenant agent management extends your existing tenant governance foundation into a centralized control plane for agent observability, governance and security. To get started, review your tenant relationships and delegated role assignments, then visit the MS Learn documentation for prerequisites, supported roles, capabilities, and licensing details.
The July 2026 updates continue to advance Microsoft Agent 365 as a comprehensive control plane for enterprise agents. With these enhancements to risk signals, cost controls, business insights, cross-platform discovery, and multi-tenant management, Agent 365 helps organizations establish consistent oversight as their agent fleet grows. Whether agents are built on Microsoft or ecosystem platforms and deployed within one tenant or many, organizations can gain a more complete view of their environment, apply governance more consistently, and make better-informed decisions about agent adoption, investment, and risk.
Next steps:
Microsoft
