Enterprise adoption of agentic AI is moving at breakneck speed, but the security risks are piling up just as fast. This week at Microsoft, Zenity—a startup specializing in securing AI agents—announced an expanded partnership with Microsoft Copilot Studio that could reshape how businesses scale AI safely.
The integration embeds Zenity’s inline attack prevention directly into Copilot Studio-built agents, giving organizations the ability to monitor, control, and stop risky actions in real time. The move addresses growing concerns around data exfiltration, prompt injection attacks, and improper handling of sensitive information—pitfalls that have dogged early deployments of generative AI in the enterprise.
Securing AI From the Inside Out
What makes this different from bolt-on monitoring is that Zenity’s controls operate within each agent, not just around them. That means whether a Copilot Studio agent is calling a CRM system, connecting to MCP servers, or processing email, Zenity can enforce guardrails at the step level—before potentially malicious instructions execute.
Think of it as giving AI copilots a security co-pilot of their own.
“By combining Microsoft Copilot Studio’s flexibility with Zenity’s security-first approach, now incorporated into inline behaviors and actions of agents, businesses can confidently deploy AI agents without sacrificing control,” said Shay Gadidi, VP of product management at Zenity.
Microsoft is framing the move as a way to unlock innovation without letting governance slip through the cracks. Shay Gurman, VP of Microsoft Copilot Studio, said the integration empowers both professional developers and business users to “innovate safely and responsibly while maintaining the security and governance that enterprises expect.”
Inline Protection, Real-Time Enforcement
Key highlights of the new integration include:
- Real-time threat disruption: Zenity can block suspicious actions based on intent and behavior before they complete, cutting off attacks mid-execution.
- Step-level policy enforcement: Every action an agent takes is monitored against enterprise rules, enabling automated compliance at scale.
- Continuous enterprise visibility: Zenity extends its platform to provide broader threat analysis and AI-specific threat reasoning across agents.
For enterprises eager to push Copilot Studio beyond IT departments, the payoff is clear: HR, marketing, finance, and operations teams can build their own agents with confidence that centralized security controls won’t be bypassed.
Why This Matters
Security is quickly emerging as the gating factor for AI agent adoption. While AI copilots promise efficiency gains across industries, enterprises remain skittish about exposing sensitive systems to unpredictable behavior. Competitors like Prompt Security and Protect AI have also entered the market with risk-monitoring solutions, but Zenity’s native integration with Microsoft Copilot Studio gives it a direct line into one of the most widely used enterprise AI platforms.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
