The leader in AI security aims to stop identity attacks on endpoints, browsers and autonomous AI agents
SentinelOne (NYSE: S), a global leader in AI-based cybersecurity, presented new solutions designed to protect human identities, also in response to the spread of AI agents and digital identities in the workplace.
Identity attacks have long been a favorite tactic of both nation states and cybercriminals. The attention of security and identity vendors has focused mainly on blocking these attacks at the root: at the authentication and authorization level. However, such attacks continue to occur, as threat actors refine tactics, techniques, and processes (TTPs) designed to bypass defenses. A hacker who accesses a system as an authorized employee and uses IT-approved tools for lateral movement and data exfiltration can cause significant damage without ever being noticed. The rise of autonomous agents that operate independently adds a new dimension to identity risks, as these agents are tasked with interacting with systems and operating without human intervention.
SentinelOne's approach is designed to stop identity attacks by following a single key principle: authorization alone is not enough. Access must be continuously validated and, if necessary, revoked during runtime. Whether on the endpoint, in the browser or within an AI workflow, operations must remain tied to real-time behavioral protections.
"The advancement of AI as an autonomous and artificial entity is increasing attack surfaces and creating new governance challenges. Identity risk no longer begins and ends with authentication, and attackers are increasingly operating within authorized workflows," said Jeff Reed, CTO of SentinelOne. “SentinelOne is at the forefront of this evolution with our native AI platform, built to correlate signals across identities, endpoints and workloads, enabling security teams to analyze behavioral intent and autonomously contain abuse by both people and unauthorized devices as it occurs.”
New identity portfolio: a single platform, a single execution structure
Modern identity attacks span browsers, endpoints, AI tools and automated workloads. To protect authorized routes, continuous validation is required in all these areas. Traditional identity platforms were designed for human users and static service accounts, not autonomous agents that execute operations and vanish in milliseconds. While human identity requires continuous verification of the user's authenticity, artificial identity requires continuous validation of intent through behaviors. Authorization alone cannot provide such validation, and the agent's behavior may deviate from its defined functionality.
The new Singularity Identity solution and SentinelOne platform architecture were designed to respond to this evolution. Built on execution, SentinelOne provides end-to-end visibility and response for human and non-human activities, such as:
- Singularity Identity provides critical context about who or what is performing an action
- Prompt Security detects abuse within the browser and AI tools
- Singularity Endpoint tests system-level behaviors
Together, these innovations enable SentinelOne to take the lead in execution-based security, providing organizations with the behavioral intelligence and autonomous control needed for modern defense.
As legitimate access increasingly becomes an attack surface and automation accelerates machine-driven activity, business resilience depends on machine-speed operations themselves. SentinelOne is transforming identities from a static gate to a dynamic behavioral assurance engine, ensuring that every action within the enterprise can be validated, trusted and, if necessary, terminated in real-time.
