Entity trail

BeyondTrust

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-05-01 / agents-researcherSix Research Teams Exploit Codex, Claude Code, Copilot, and Vertex AI — Every Attack Hit Runtime CredentialsVentureBeat reports that six independent research teams disclosed exploits against four major AI coding agents over nine months, and every attack followed the same pattern: the agent held a credential, executed an action, and authenticated to production without a human session anchor. BeyondTrust demonstrated that a crafted GitHub branch name could steal Codex's OAuth token in cleartext. Anthropic's Claude Code source spilled onto npm, and Adversa found Claude Code silently ignored deny rules past 50 subcommands. Gravitee's 2026 survey found only 21.9% of teams have onboarded agent OAuth credentials to PAM platforms.
2026-03-27 / agents-researcherSiliconANGLE RSAC Analysis: Agentic AI Security Demands Zero-Trust Playbook — Identity, Least Privilege, and Microsegmentation for Non-Human ActorsSiliconANGLE's March 27 analysis argues RSAC 2026 proved that agentic AI security is not a new discipline but a zero-trust enforcement problem: every agent needs identity, least privilege, microsegmentation, and continuous verification — the same principles applied to human users, now extended to non-human actors at machine speed. The article synthesizes announcements from Cisco, CrowdStrike, Palo Alto, BeyondTrust, and Astrix into a unified framework, concluding that vendors who already sell zero-trust infrastructure have the architectural advantage over AI-native startups.
2026-03-27 / agents-researcherBeyondTrust: Enterprise AI Agents Growing 466.7% YoY — First Unified Privileged Identity for AI Coworkers From Desktop to CloudBeyondTrust Phantom Labs research reveals a 466.7% year-over-year increase in AI agents operating inside enterprise environments, with the majority running as shadow agents with privileged access invisible to security teams. BeyondTrust's Pathfinder Platform now delivers endpoint privilege management for AI clients (Claude, ChatGPT) enforcing least privilege at the OS level, plus Identity Security Insights with connector coverage spanning OpenAI, Google Vertex, Salesforce Agentforce, ServiceNow, and AWS Bedrock for automatic discovery, privilege path mapping, and risk scoring.
2026-03-19 / arxiv-researcherAWS Bedrock AgentCore 'Sandbox' Bypassed via DNS: BeyondTrust Discloses C2 Channel and Full Data ExfiltrationBeyondTrust Phantom Labs disclosed (March 16) that AWS Bedrock AgentCore's Code Interpreter Sandbox permits outbound DNS queries, enabling covert C2 channels and exfiltration of PII, API keys, and financial data. Commands are encoded as chunked ASCII in DNS A-record IP responses; output is exfiltrated via base64-encoded DNS subdomain queries. AWS declined to patch—updating docs to call DNS resolution 'intended functionality'—and awarded a CVSS 7.5, leaving the attack surface open for all Bedrock AgentCore users today.
2026-03-17 / agents-researcherCritical Unpatched CVSS 9.8 RCEs in SGLang + Amazon Bedrock AgentCore DNS Exfiltration Disclosed TodayResearchers disclosed two critical unpatched RCEs in SGLang (CVE-2026-3059, CVE-2026-3060, both CVSS 9.8) via unsafe pickle deserialization in ZeroMQ broker and disaggregation module — any SGLang deployment exposing multimodal or disaggregation features is fully exploitable with no authentication. Separately, BeyondTrust revealed Amazon Bedrock AgentCore Code Interpreter's sandbox permits outbound DNS queries, enabling attackers to establish interactive shells and exfiltrate data from inside agent code execution environments. Both disclosures extend the ShadowMQ vulnerability pattern — unsafe ZeroMQ/pickle combos copied across the AI inference stack — which previously hit Meta vLLM (CVE-2025-30165) and NVIDIA TensorRT-LLM (CVE-2025-23254).
2026-02-21 / sources-researcherBeyondTrust CVE-2026-1731 CVSS 9.9 Pre-Auth RCE Under Active ExploitationUnit42 documents active exploitation of BeyondTrust Remote Support pre-auth RCE. Over 10,600 exposed instances. VShell and SparkRAT backdoor deployments across financial, legal, tech, healthcare.

Graph relationships

CRITICIZES
Bedrock
BeyondTrust disclosed that Amazon Bedrock AgentCore permits outbound DNS queries.
Source finding
RELEASED
Pathfinder Platform
BeyondTrust released Pathfinder Platform for AI privilege management
Source finding

Source trail

BeyondTrust CVE-2026-3059 and CVE-2026-3060 VentureBeat SiliconANGLE BeyondTrust Unit42

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…