Silverfort Adds Runtime Identity Controls to Microsoft Copilot Studio Agents

Quick Facts

• Silverfort now evaluates every access request as Copilot Studio agents begin actions, blocking unauthorized activities before execution
• The integration addresses identity security gaps as AI agents exceed privileges and interact with systems in unpredictable ways
• Silverfort has raised $222 million total, including a $116 million Series D round led by Brighton Park Capital in January 2024

Silverfort integrated runtime identity controls with Microsoft Copilot Studio agents to address security gaps as artificial intelligence systems operate autonomously across enterprise applications.

The security platform now evaluates every access request just as Copilot agents begin to act. This real-time enforcement blocks unauthorized actions before execution, according to the company.

Traditional identity and access management systems were designed for humans and service accounts, not autonomous systems that act continuously. AI agents can exceed privileges, expose sensitive data, or interact with systems in ways security teams cannot predict.

Runtime Protection Model

Silverfort emphasizes runtime enforcement instead of relying on preconfigured permissions or after-the-fact monitoring. The platform evaluates context continuously, including the identity behind the agent, requested actions, risk signals, and current policy state.

This approach matters because AI agents operate quickly and autonomously. Static permissions can become overprivileged while retrospective logging may spot abuse only after damage occurs.

The integration treats AI agents as first-class identities requiring the same security principles as human users: least-privilege access, contextual authorization, continuous monitoring, and auditable accountability. Silverfort connects every agent back to human owners while evaluating runtime activity against enterprise identity policies.

Market Context and Funding

Silverfort has raised $222 million in total funding. The company closed a $116 million Series D round led by Brighton Park Capital in January 2024.

The company positions itself as the identity security layer filling gaps left by platforms like Okta and Azure AD. These platforms can enforce multi-factor authentication for modern applications but cannot extend controls to legacy protocols including NTLM, Kerberos, LDAP, and RDP.

Recent research shows 54% of executives cite AI-enhanced threats as their top concern for 2026. Meanwhile, 82% say legacy infrastructure actively creates identity security risk.

Microsoft Partnership

Silverfort signed a partnership with Microsoft in March 2021 to enable organizations to consolidate identity security on Microsoft Entra ID. The platform natively integrates with Microsoft Sentinel, feeding identity telemetry into existing security information and event management systems.

Microsoft Copilot Studio is accelerating AI agent adoption by helping organizations build and deploy systems at scale. The platform recently added real-time voice agents, now generally available in North America through Dynamics 365 Contact Center.

This integration positions both companies in the emerging agentic AI security market as organizations increasingly deploy AI agents for business-critical processes.

Read more: Silverfort brings runtime identity controls to Microsoft Copilot Studio agents