Check Point partners with Amazon Bedrock AgentCore to extend security tooling to enterprise AI agents

Check Point Software Technologies said it is bringing its AI-focused security capabilities into Amazon Web Services’ Bedrock AgentCore, a step aimed at addressing security and governance needs for enterprise AI agents and workflows. The announcement, reported by Yahoo Finance on June 28, positions the integration as a way to protect organizations using Bedrock-based agents that can perform tasks and interact with business systems in automated or semi-automated ways.

The collaboration centers on securing “enterprise AI workloads,” according to the coverage. While the report does not spell out specific technical measures in detail, it frames the Bedrock AgentCore integration as a means to extend Check Point’s security controls to a growing class of AI deployments that are increasingly interactive and action-oriented rather than limited to static chatbot responses.

For Amazon, Bedrock AgentCore is part of AWS’s broader effort to help customers build and run AI agents using managed services. AgentCore is positioned within AWS as a foundation layer for agent development, with Bedrock serving as the broader platform for deploying and managing foundation-model based applications. In this context, security partners like Check Point are seeking a role in the shared responsibility stack, where enterprises increasingly ask for guardrails around data access, output handling, and overall agent behavior.

Check Point, meanwhile, has been working to adapt its security portfolio to AI use cases that require more than traditional perimeter defense. The company’s stated focus in this partnership is on AI security for enterprise deployments, suggesting its tooling is intended to detect, prevent, or manage risks that arise when AI systems are granted access to business resources and when agent workflows evolve over time.

The Yahoo Finance post does not provide additional operational specifics such as which Check Point products are involved, what policies or detection categories are covered, or whether the integration is generally available or limited to certain early access customers. It also does not list performance benchmarks, deployment timelines, or how the integration works at the data and API layer.

A related caveat is that the report does not quantify the partnership’s commercial impact. There are no disclosed revenue targets, customer counts, or contract values. As with many technology integrations announced through media channels, the practical value for buyers will depend on implementation details that are not included in the brief announcement coverage.

Still, the direction of travel is clear for both companies. As enterprises move from isolated AI experiments to agent-based workflows, security vendors are looking to attach protections closer to the AI platform rather than only around the network or endpoints. For buyers, that typically means evaluating whether security controls can cover the full lifecycle of AI interactions, including how agents are allowed to read from and write to systems of record.

Over the near term, organizations watching this space may want to look for Check Point’s documentation on the integration scope, availability, and supported Bedrock AgentCore features, as well as independent verification of how the security controls map to common AI risks such as prompt manipulation, data leakage, and unsafe or policy-violating actions. The immediate next checkpoint is whether the companies publish deeper technical materials beyond the headline announcement.