As the number of AI agents deployed across enterprises continues to soar, concerns about their security and governance have intensified. Recent updates to Okta's AI Agents platform aim to tackle these challenges by integrating with Amazon's Bedrock AgentCore. This integration supports a range of identity providers and enhances organizations' ability to manage and govern agent access to resources more effectively.
Ely Kahn, Chief Product Officer at Okta, stressed the urgency of the situation, stating, "Security and IT leaders need a better way to understand where their agents are, what they can connect to, and what they can do." As enterprises increasingly deploy AI agents in various environments, it has become clear that a single-vendor ecosystem is no longer adequate for effective management. Okta's platform is designed to be neutral, providing comprehensive lifecycle management—from discovery and onboarding to ongoing governance and protection.
The integration with Amazon Bedrock AgentCore enables enterprises to create a governed identity layer tailored for their AI agents without extensive infrastructure overhauls. Shakeel Khan, RVP and Country Manager for Okta India, highlighted a significant gap in governance, noting that 90% of enterprise agents are over-permissioned, with more than half accessing sensitive information. This trend raises concerns about the visibility and control organizations have over their agent deployments.
Addressing Governance Challenges
The rapid proliferation of AI agents presents unique governance challenges for enterprises. Gartner predicts that by 2028, an average Fortune 500 company will utilize over 150,000 agents, raising the stakes for effective management. Many organizations struggle with over-permissioned agents, increasing the risk of sensitive data exposure. Okta's enhancements aim to bridge this visibility gap, offering tools that help enterprises manage agent permissions effectively.
The integration allows for essential functionalities such as AI Agent Discovery, which identifies agents operating within an organization's ecosystem by monitoring new OAuth consent grants. This capability is crucial for maintaining oversight in environments where agents may be deployed without adequate governance frameworks. The AI Agent Import feature simplifies the process of importing agents from AgentCore into Okta's platform, enabling swift governance measures within minutes.
https://www.youtube.com/watch?v=lln79Qt8vH8
Unified Identity Layer Across Ecosystems
The current environment of AI deployment is characterized by rapid changes in model leadership and shifting platform adoption patterns. Organizations that engage with multiple AI frameworks require a unified identity layer that operates seamlessly across these ecosystems. Traditional point solutions, which often secure isolated components of the agent lifecycle, fail to deliver comprehensive security.
Okta's approach seeks to address this gap by providing a vendor-neutral platform that encompasses the entire agent lifecycle. This not only enhances security but also ensures consistent access policies across various identity platforms. As enterprises navigate the complexities of AI agent governance, Okta's solutions emerge as a critical resource for achieving responsible AI deployment at scale.
In a landscape where the deployment and governance of AI agents are evolving rapidly, Okta's integration with AWS represents a strategic move. By facilitating a governed identity layer for AI agents, Okta helps enterprises navigate the complexities of AI security more effectively, ensuring they scale with the necessary oversight and control.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
