Box expands enterprise AI governance with new agent security featuresox

Box has announced new security capabilities designed to give organizations greater control over AI agents working with enterprise content. With new agent guardrails, third-party agent activity oversight, prompt injection detection, agent classification-based access policies, and more, customers will be able to extend Box’s security controls to both Box Agents and third-party agents, such as Claude, ChatGPT, and Gemini.

These new capabilities expand Box’s trusted governance framework to agentic workflows, enabling organizations to confidently deploy and scale AI agents while maintaining oversight of how enterprise data is accessed and used.

“83 percent of organizations are already experimenting with AI agents across their most critical tasks,” said Manoj Asnani, VP of AI Security, Privacy, Compliance & Governance Products at Box.

“As these agentic workflows become more deeply embedded in the enterprise, it’s critical to create the proper security controls to ensure agents have what they need to function effectively, without accessing, modifying, or exposing content beyond the scope of its intended task. Box already provides a safe environment for organizations to apply AI to their most critical business knowledge, and with these new controls, we are creating a standard for deploying agents of any kind securely and at scale.”

“As we rapidly advance our utilization of AI agents, we expect Box—which has consistently led the development of security management capabilities for secure collaboration—to provide the administrative features needed to safely leverage this new era of AI,” said Tatsutoshi Murata, Head of IT Strategy Department at Nomura Research Institute.

“In particular, we’ve found it extremely reassuring that Box offers multi-vendor support, allowing us to flexibly switch between AI models, while providing security management capabilities that span prevention, detection, and response. With a protective layer that appropriately manages AI agent access to content, we’re confident our critical content will remain protected as we expand our use of AI,” Murata continued.

Introducing new security capabilities for protecting enterprise content

Box’s 2026 State of Enterprise AI report found that security and privacy are the leading obstacles to deploying AI agents at scale. Among IT leaders surveyed, 90% identified security, regulatory, and trust concerns as the biggest barrier to granting AI agents access to enterprise content.

Box’s new security and governance capabilities address this concern by being built directly into the platform where content already lives. By providing protections at the content layer, these controls help ensure that every agent action is intentional, permissioned, and auditable.

Whether agents are built in Box or connected through third-party platforms, security is applied at every touchpoint. This allows organizations to move from limited pilots to production-scale AI deployments without compromising on governance.

Box’s new security and governance capabilities do not require additional tools to deploy or manage, and include:

• Agent guardrails that define exactly what custom Box AI agents can and cannot do based on content sensitivity and policy, including enforcing label-based access controls, requiring approval for deletion actions, and disabling external sharing to help protect sensitive information;
• Prompt injection detection capabilities that validate every input before it reaches the model by detecting known prompt injection patterns at the content layer and enabling organizations to log, alert on, or block suspicious attempts;
• MCP guardrails that let admins control what external AI agents connected through the Box MCP Server are allowed to do, with scoped permissions such as allowing file creation only in approved folders, blocking external sharing, and permitting content moves only to specific folders;
• Classification-based access policies for both external and custom Box AI agents that enable organizations to exclude content with specified classifications from being read, searched, or accessed;
• Agent activity oversight capabilities that provide visibility into external AI agent activity on customer content and configure threshold-based alerts, enabling organizations to quickly detect and respond to suspicious behavior;
• Agent audit trails and session governance that retain, audit, and provide compliance-ready records for every agent session with full session context, including retention policies and legal holds;
• Human-in-the-loop control features that require human approvals before agents execute sensitive or high-impact actions.

With these new security and governance capabilities, organizations across financial services, healthcare, legal, insurance, and more can confidently deploy AI agents.

• Financial services firms can safeguard sensitive M&A analysis and trade information by enforcing agent-level access controls and catching unexpected agent behavior in real time;
• Healthcare organizations can secure patient data and proprietary research by enforcing classification-based access controls and retaining a complete, compliance-ready audit trail of every agent action;
• Law firms can govern AI across contract analysis and discovery workflows by enforcing policy-driven controls on what agents can access and act on;
• Insurance companies can protect sensitive policyholder data and claims records by validating every agent input for prompt-injection attempts and excluding classified records from being accessed, ensuring only authorized AI agents touch confidential claims and regulatory information across claims processing and reporting workflows.

“Box’s new security and governance controls address the primary barriers of privacy and unauthorized access directly where the data lives. By embedding guardrails, prompt injection detection, and human-in-the-loop oversight into the platform, Box is establishing a vital trust standard that allows enterprises to confidently scale both native and third-party AI agents across their most sensitive content,” said Amy Machado, Senior Research Director, Content and Knowledge Management Strategies, IDC.