Reco, the software-as-a-service (SaaS) security platform, has launched a groundbreaking feature aimed at enhancing the security of artificial intelligence (AI) agents within enterprise environments. The new capability, known as Reco AI Agent Security, provides security and risk teams with unprecedented visibility and control over AI agents operating across various SaaS platforms, including Copilot, ChatGPT, and Salesforce Agentforce. This feature is available immediately as part of Reco’s unified security solution.
As organizations increasingly integrate AI into their operations, they face dual challenges of AI sprawl and agent sprawl. AI sprawl refers to the rapid proliferation of AI services, while agent sprawl is characterized by the unchecked growth of autonomous AI agents capable of accessing sensitive data and executing actions without direct oversight. These challenges complicate the responsibilities of security teams, making effective governance and oversight critical.
Reco AI Agent Security addresses a significant gap in existing security frameworks. Traditionally, SaaS security tools have focused on user identities and application visibility, leaving a blind spot regarding autonomous AI agents. These agents, which can operate independently, often possess extensive permissions, interact across multiple SaaS platforms, and function outside the purview of IT and security teams. The new capability from Reco aims to fill this void by inventorying every AI agent, mapping their access, permissions, and risk posture, thereby enabling security teams to make informed decisions regarding management and control.
The key functionalities of Reco AI Agent Security include a comprehensive inventory of AI agents, which automatically detects agents across various platforms and integrations. It also offers detailed mapping of access and permissions, allowing teams to understand which identities and data the agents can access. Additionally, it identifies and prioritizes risks associated with agents that have excessive permissions or exposed credentials. Governance controls allow security teams to enforce policies and block unauthorized agents efficiently. The feature also utilizes a knowledge graph to correlate identities, applications, and security contexts, offering actionable insights for remediation. This integration facilitates immediate responses to high-risk agents, including revoking permissions or disabling unauthorized agents.
In contrast to other recent trends in AI security that primarily focus on posture management and runtime protections, Reco’s solution uniquely addresses the risks presented by autonomous AI agents embedded in SaaS workflows. While other security providers are beginning to integrate AI into their threat management processes, Reco stands out by securing applications, identities, and now AI agents within a unified governance framework. This approach is particularly significant as enterprises face the complexities of modern risk management in an increasingly interconnected landscape.
Ofer Klein, CEO and Co-Founder of Reco, emphasized the growing complexity of modern enterprise environments. “Enterprises today don’t just have hundreds of connected SaaS apps — they have thousands of connected AI agents operating in the background,” Klein stated. “Unlike traditional SaaS plugins, AI agents can act autonomously and span identity, data, and systems, exponentially increasing risk when misconfigured or unmanaged.” The introduction of Reco AI Agent Security is positioned as a necessary measure for security teams striving to maintain control and oversight.
The new capability is now available to both existing and new Reco customers, launching with built-in integrations for popular platforms such as Copilot, ChatGPT Enterprise, Salesforce Agentforce, Make, and n8n. Reco plans to support additional agents and custom automation tools on a continuous delivery cycle, ensuring that organizations remain equipped to manage the evolving landscape of AI and SaaS security.
See also
Anthropic’s Claims of AI-Driven Cyberattacks Raise Industry Skepticism
Anthropic Reports AI-Driven Cyberattack Linked to Chinese Espionage
Quantum Computing Threatens Current Cryptography, Experts Seek Solutions
Anthropic’s Claude AI exploited in significant cyber-espionage operation
AI Poisoning Attacks Surge 40%: Businesses Face Growing Cybersecurity Risks
