BeyondTrust introduces tools for securing AI agents
Identity security company BeyondTrust has introduced additional capabilities to its Pathfinder Platform aimed at providing a unified approach to securing AI agents and autonomous AI workloads.
The expanded capabilities include an endpoint privilege management tool capable of enforcing least privilege and application control for AI clients operating in enterprise endpoints, as well as AI agent discovery and risk analysis capabilities providing agentic AI discovery, classification and posture auditing.
The tools support major enterprise AI platforms including OpenAI’s Admin, Projects and ChatGPT Enterprise, Google Vertex AI and Discovery Engine, Salesforce Agentforce, ServiceNow AI agents, and AWS Bedrock.
Other additional toolsets include a secrets management solution for autonomous agents that can enforce just-in-time access to the secrets and API keys behind agentic workloads, and a tool designed to manage the rapid acceleration of AI agents across enterprise environments.
BeyondTrust CTO Mark Maiffret said AI agents are increasingly being deployed across the enterprise, but many organisations lack visibility into how many are operating in their environments and what data these agents can access.
“Agentic AI is not an isolated problem. It’s a subset of the broader non-human identity landscape,” he said. “Organisations cannot secure agentic identities in a silo. These agents are interconnected with human identities, machine accounts, secrets and entitlements across every environment.
“The question security teams should be asking isn’t ‘Do we have AI agents?’ You do. The question is: what can they access, what secrets are they using, and what happens if one gets compromised? A single AI agent’s blast radius can span your identity providers, cloud infrastructure, SaaS platforms and on-prem directories all at once. Pathfinder maps those cross-domain privilege paths so you can see exactly how a compromised agent could escalate access across your entire environment.”
CrowdStrike and NVIDIA develop secure-by-design AI blueprint
CrowdStrike has arranged to integrate its Falcon security platform into the open-source NVIDIA...
Unit42 discloses severe flaw in Google's Gemini
Researchers from Palo Alto Networks' Unit 42 have detailed their discovery of a...
SentinelOne unveils identity security portfolio
SentinelOne has developed its security platform with capabilities aimed at safeguarding...
