Reco to Present Sessions on AI Agent Security Exposure at Black Hat USA and DEF CON 34

Reco, an AI and agent ecosystem security company, is scheduled to present two sessions at Black Hat USA and DEF CON 34 focused on risks that appear when AI agents are embedded into enterprise applications and workflows.

The presentations address how agents, which are not limited to prompts, models or standalone AI tools, can create an attack surface that exposes applications, identity systems, customer portals and business workflows to security risks. The company’s sessions also cover how agent behavior can involve inherited permissions and OAuth grants that trigger actions and expose data through trusted business systems.

At Black Hat USA, Shir Grinfeld, Head of Product Growth at Reco, will present a three-phase framework for reducing the exposure window created by AI agents. The framework includes identifying misconfigured apps, unenforced SSO policies and over-privileged tokens; detecting and responding to machine-speed OAuth probing during an incident; and preserving access trails for post-incident remediation and executive reporting.

At DEF CON 34, Nitay Bachrach, a security researcher at Reco, and Cynthia Ardman, a threat detection engineer at Reco, will co-lead “Salesforce Apex Predator: Breaking Salesforce Sites,” a hands-on workshop focused on Salesforce Experience Sites. The workshop will cover Aura and LWR frameworks, custom Apex controllers running in system mode, SOQL injection, unauthenticated route enumeration and LWRed, described as a new open-source scanner for LWR sites. “Reco executives and researchers will be available at Black Hat and DEF CON to discuss how AI agents are creating unrecognized and ungoverned risk across enterprise ecosystems, and what organizations can do to protect against business disruption and data leakage,” the company said.

Provided by Globe Newswire on behalf of Recorded Future. Click to read original content. The original article was written by Decision Insights Editorial.