News for today:
- [BeyondTrust] Black Hat USA 2026 releases. At Black Hat this week, BeyondTrust is showcasing new native capabilities in Pathfinder and adding new context and telemetry enrichments. PathfinderAI & MCP Server, AI Agent Security, NHI Governance, and Workload Credentials are the first native Pathfinder capabilities designed to help organizations discover, prioritize, govern, and protect privileged access wherever it exists. Together, these capabilities bring human and non-human identities, including AI agents and workloads, onto a single platform that combines identity visibility, intelligence, and protection to help organizations uncover, understand, and reduce the standing privilege attackers depend on. BeyondTrust
- [Arctic Wolf] Agentic SOC – numbers are better than before. Arctic Wolf updated the performance metrics of its Agentic SOC offering, and reinforced its wider purview around scale, value, trust, and speed (versus other offerings which they assert focus on detection speed only). The Aurora Agentic SOC resolves many investigations in as little as 12 seconds while delivering actionable remediation guidance that helps customers respond with confidence. Those efficiencies are translating into measurable results, with customers resolving cases 26% faster year over year and spending less time interpreting alerts and more time taking action. Arctic Wolf
- [Varonis] Intent-based access control for agents. Varonis added intent-based access control for agents to its Atlas platform, for enforcing safeguards on AI agent behaviors. Agent IBAC checks whether an agent’s reasoning, tool use, and data access are consistent with its assigned instructions, and alerts on or blocks actions that do not align. When an agent crosses the line, Atlas can quarantine the identity behind it and block everything that follows for a defined window. Varonis
- [KnowBe4] Claude support. KnowBe4 added support for Claude to its Agent Risk Manager (was previously only for Microsoft Copilot), bringing agent governance to Claude agents. While employees use Claude-powered agents to automate complex tasks, analyze data, and connect with third-party apps, these AI identities often operate outside normal security controls. Agent Risk Manager closes this blind spot with a simple to add security layer that monitors and governs agent execution without modifying the underlying model, extending the same risk-based approach KnowBe4 has applied to human behavior for 15 years to the agentic workforce. KnowBe4
- [OPSWAT] Cybersecurity docuseries almost here. OPSWAT announced that its new docuseries will be available on August 8; the overall focus is cybersecurity, and the first episode is on firewalls. Presented and commissioned by OPSWAT, and produced by Cyber King Productions, the debut episode, “Breaking the Firewall,” tackles one of the industry’s common and dangerous assumptions: that firewalls alone are sufficient to protect critical assets. Kari Byron of “MythBusters” hosts and guides viewers through hands-on experiments, cinematic demonstrations (including a physical firewall experiment involving a flamethrower and explosions), and expert insights. She explores how attackers can bypass traditional firewall and how technologies such as data diodes can add a deterministic hardware enforced layer of protection for critical environments which is unbreakable. OPSWAT
- [Qualys] InstaScan for continuous exposure visibility. Qualys released a new continuous exposure visibility capability in its Enterprise TruRisk Management platform – for faster analysis of which vulnerability announcements impact a given organization. InstaScan introduces scanless scanning, an innovative, autonomous vulnerability management capability within Qualys ETM. Powered by Agent Insta, a cyber risk agent leveraging AI to continuously monitor newly published vendor security advisories and threat intelligence from Qualys and trusted third-party sources, Agent Insta correlates emerging vulnerabilities with an organization’s live inventory, telemetry and threat intelligence. It automatically identifies impacted assets and surfaces trusted detections within minutes. Qualys
- [Horizon3] $250m Series E funding. Horizon3 announced Series E funding of $250 million, at a valuation of $2 billion. Funds will be used for scaling go-to-market operations, expanding internationally, and accelerating the product roadmap. The oversubscribed round was co-led by existing investors NightDragon and NEA, with participation from seven new investors and five returning backers. The capital underscores accelerating global demand for safe, autonomous security validation as AI-driven cyberattacks escalate. Horizon3