SentinelOne (NYSE: S) announced an upcoming integration of its Prompt Security offering with Amazon Bedrock AgentCore to deliver runtime guardrails for organizations deploying AI agents at enterprise scale. The integration aims to extend existing security controls into agentic workflows, empowering enterprises to secure and scale their AI operations on AWS. By combining SentinelOne's detection capabilities with AgentCore's policy engine, the solution addresses the security risks associated with large-scale AI agent deployments.

Amazon Bedrock AgentCore serves as a platform to build, connect, and optimize agents at scale. Its policy engine enforces real-time, deterministic controls at the gateway across all agent traffic, including agent-to-tool, agent-to-LLM, and agent-to-agent communications. SentinelOne's Prompt Security will provide detection signals covering prompt injection detection, PII exposure, tool-use validation, LLM response monitoring, and data leakage prevention. These signals feed into the AgentCore policy engine, which acts on them at the gateway, outside the agent's reasoning loop.

The architecture ensures that while detection may be probabilistic, enforcement remains deterministic. This allows enterprise teams to apply SentinelOne's detection signals alongside automated enforcement consistently across all agent activity. Customers gain consistent allow-or-deny decisions on every agent action, enabling the adoption of agentic AI with the confidence that existing SentinelOne security controls apply automatically throughout every workflow.

"Agentic AI has become foundational to how enterprises operate," said Melissa K. Smith, SVP of Global Strategic Partnerships and Initiatives at SentinelOne. "Every organization building and running AI agents on AWS needs the same security controls they rely on across the rest of their infrastructure. This upcoming integration applies those guardrails at the AgentCore gateway, where all agent traffic is evaluated, giving every enterprise the confidence to deploy AI agents knowing their existing SentinelOne policies apply automatically, from day one."

The integration will be available in Amazon Bedrock AgentCore later this year. Customers with existing Prompt Security licenses will be able to apply them through a Bring Your Own License path. As enterprise AI agent deployments scale, consumption is expected to scale accordingly.

Key Features of the Integration

SentinelOne has opened Purple AI Agentic Investigation to all customers, introducing a zero-click, autonomous capability that detects, investigates, and responds to threats within the existing Singularity Platform workflows. The company also introduced Singularity Credits, a unified currency for running AI-powered work across the platform. This launch addresses the investigation capacity constraints faced by security teams, where alerts often queue for attention and verdicts depend on analyst availability.

The new capability is designed to function as a force multiplier for analysts by autonomously initiating investigations when a threat crosses a defined threshold. Purple AI collects evidence, correlates telemetry, and builds an attack timeline, allowing analysts to begin at the verdict rather than the initial alert. The system operates on telemetry already present in the platform across endpoints, identity, cloud, and third-party security data, requiring no additional configuration or data deployment.

Key Features of Purple AI Agentic Investigation

Purple AI is built to integrate seamlessly with the Singularity Platform, offering governed autonomy with full auditability. Key features include:

• Seamless Integration: Built into the Singularity Platform, requiring zero configuration and no data leaves the environment.
• Analyst Force Multiplier: Automates investigation work to scale capacity without increasing headcount, freeing analysts for judgment and response decisions.
• Auditable Autonomy: Every verdict includes a complete evidence chain, with adjustable human-in-the-loop settings based on SOC maturity.
• Advanced Reasoning: Utilizes a multi-model approach combining Anthropic's Claude, OpenAI's GPT, and SentinelOne's proprietary "Ultraviolet" models to compress investigations.

Singularity Credits and Availability

Singularity Credits serve as the flexible currency for AI-powered work within the platform. To facilitate adoption, SentinelOne is granting customers a complimentary allotment of Credits to trial the Agentic Investigation capability. The trial is currently available in Singularity Platform consoles for new and existing customers, requiring no payment method during the period. Post-trial, Credits can be purchased through partners, direct billing, and eCommerce.

Strategic Vision for the Agentic SOC

This advancement supports SentinelOne's vision of the agentic Security Operations Center (ISOC), where frontier-AI reasoning amplifies human defenders. By operating natively on diverse telemetry sources, Purple AI aims to transform the Singularity Platform into an integrated system that autonomously detects, triages, and stops threats at machine speed.