SalePoint is partnering with Amazon Web Services (AWS) to expand its identity security framework to include AI agents. This move comes as the issue of consistently managing who is granted what permissions has emerged as a core security challenge, driven by the increasing number of environments where humans, applications, machines, and AI agents work together within enterprise systems.

SalePoint announced on the 26th that it has signed a multi-year strategic cooperation agreement with AWS. This agreement expands the scope of existing cooperation and focuses on managing human and non-human identities within a single governance layer in the AWS environment. Through this, the two companies plan to jointly develop a security and governance framework suitable for the expansion of agentic AI.

The background of this collaboration lies in the proliferation of AI agents. While AI agents perform tasks on behalf of users or systems, they simultaneously operate as new non-human identities. Concerns have been raised regarding the increase in accounts and privileges during this process, as well as the potential for unmanaged access paths to become attack surfaces. SalePoint and AWS envision applying the same policies to all identities in this environment and maintaining control over workloads across AWS.

The cooperation direction proposed by both companies focuses on managing the entire lifecycle—from account creation to authentication, authorization, and revocation—within a single framework. This includes continuously applying the principle of least privilege by leveraging AWS CloudTrail-based usage data, as well as an Identity Graph that provides an integrated understanding of access relationships among workloads, federated identities, services, and data. The plan also includes establishing a policy enforcement system that automatically revokes access in the event of risks or role changes.

Technically, a structure was proposed in which SalePoint integrates with AWS AgentCore to detect AI agents in the environment and manage them as a single identity. This enables enterprises to view human and agent accounts together on a single management screen, allowing them to track human-agent connections, review access, optimize permissions, and enforce policies. In the future, features are also being considered to support creating accounts on behalf of agents, as well as requesting and approving new access permissions.

Truck manufacturer PACCAR was cited as a customer case. PACCAR explained that by utilizing SalePoint Identity Security Cloud, which operates on AWS infrastructure, they transitioned identity management into an automated process, which helped ensure regulatory compliance and scalability. Moving forward, both companies plan to expand the adoption of the relevant solution by engaging in sales and marketing collaboration in addition to technological cooperation.

This contract demonstrates the trend of expanding identity security from human-centric management to non-human entities in an enterprise environment where the pace of AI adoption is accelerating. As AI agents enter actual business systems, it is becoming increasingly clear that security must be designed to include authority and control structures, rather than just model performance.