News for today:
- [Datadog] AI agent for SOC investigations delivers 98% reduction. Datadog released its Bits AI Security Analyst, an AI agent for SOC investigations, as part of the Datadog Cloud SIEM. Claims a reduction in mean-time-to-resolution of up to 98%. Bits AI Security Analyst [pairs] the expertise of a senior SOC analyst with machine scale and speed, enabling investigation analysis across a breadth and volume of data sources that would be unachievable by a human, while still delivering high-accuracy verdicts backed by real-world context. This allows analysts to scale their investigation expertise so they can focus more time on high-impact defense priorities. Available immediately. Datadog
- [Rubrik] Real-time AI governance engine for autonomous agents. Rubrik unveiled the Semantic AI Governance Engine (SAGE), giving intent-driven governance for AI agents and agent behavior. Enterprise AI deployment is stalling at a governance bottleneck, as legacy systems rely on deterministic rules that cannot comprehend natural language nor adapt to dynamic and unforeseen actions taken by agents. Rubrik SAGE solves the bottleneck by using Rubrik’s custom Small Language Model (SLM) to interpret the semantic meaning of policies, providing a real-time command center for agentic operations. Among other capabilities: SAGE translates natural language instructions (e.g., “Do not give financial advice”) into machine logic, recognizing context that static filters miss. Rubrik
- [Palo Alto Networks] Updates to the Prisma Browser. Palo Alto Networks updated its Prisma Browser for secure agentic AI interactions. New capabilities include: [1] Automatically discovers user AI activity and enforces content-aware boundaries to keep agents within their intended scope. Prisma Browser prevents sensitive data from leaking to unmanaged or public AI tools during automated tasks. [2] Identifies and blocks prompt injection attacks—including malicious instructions hidden within websites designed to hijack AI agents—keeping automated workflows on track and preventing agents from being manipulated into unauthorized actions. Palo Alto Networks. There’s a version for small businesses, too, called the Prisma Browser for Business. Palo Alto Networks
- [Netwrix] Visibility and governance over AI agent / assistant access to sensitive data. Netwrix updated its 1Secure platform with new capabilities to secure AI access to sensitive data, via a combination of identity, data discovery, and access monitoring. Among other things, new capabilities in Netwrix Access Analyzer provide deeper insight into how identities access sensitive data across hybrid environments. Security teams can identify excessive permissions, hidden access paths, and risky identity relationships that could expose sensitive information to AI systems, automation tools, or unauthorized users. There’s more new feature goodness across complementary products – Auditor, Threat Prevention, Endpoint Protector, and Threat Manager. Netwrix
- [Drata] New unified company brand. Drata announced that its dropping the SafeBase brand name from Trust Center, unifying around Drata for consistency. SafeBase was acquired last year. For our customers, continuity is paramount. The Trust Center remains a core component of the Drata platform. AI Questionnaire Assistance continues as a central capability. Existing features, integrations, contracts, pricing, URLs, and security standards remain unchanged. Ownership has been under Drata since the acquisition, and this brand unification does not alter how data is stored, handled, or supported. Drata
- [KnowBe4] Reporting phishing in Microsoft Teams. KnowBe4 introduced a button for Microsoft Teams for reporting phishing in chats and channels. The Phish Alert Button is a simple, yet powerful, one-click reporting tool designed to turn end users into a human network of security ‘sensors’. When an employee encounters a message that appears suspicious, unexpected, or potentially malicious, they can use the PAB to report it immediately. Once clicked, the tool automatically forwards the message to the organization’s designated security inbox and incident response team for analysis. KnowBe4