News for today (with a bit of catchup):
- [RadarFirst] AI Incident Management. RadarFirst GA’d its new solution for investigating, managing, and documenting AI-related incidents. Radar AI Incident Management provides a centralized, structured process for capturing the facts and evidence needed to evaluate potential harm across categories such as bias and discrimination, harmful outputs, privacy, safety, and policy violations. By bringing together incident context — including relevant information about the AI system and its role — the solution helps teams apply defined harm thresholds, make consistent determinations, and document the decisions, approvals, and actions that follow. RadarFirst
- [Rapid7] More intelligence. Rapid7 launched a new unified engine to deliver proactive prevention for clients from threat data, not just reactive monitoring; it’s the replacement for the legacy Rapid7 Labs offering. Building on Rapid7 Labs’ 13-year legacy, Rapid7 Intelligence turns attacker behavior into scalable protection faster than traditional vendors reliant on static indicators. With AI compressing exploitation windows to minutes—and 8,539 critical vulnerabilities tracked in Q2 alone—attackers are leveraging automation to scale phishing, reconnaissance, and script development. Rapid7 Intelligence monitors these threats in real time to deliver actionable, expert-validated insights that keep defenders ahead. Rapid7
- [Sophos] CISO in an (AI) box. Sophos launched an alternative to a CISO – an AI-enabled solution that creates informed cybersecurity strategy from underlying security data. Sophos CISO Advantage …. builds a security assessment unique to each organization’s environment and threat profile, maps controls against frameworks including NIST CSF, CIS v8, Cyber Essentials Plus, and NCSC CAF, and turns the results into a prioritized, budget-aligned roadmap of what to fix first, what it costs, and why it matters to the business. Because it is part of Sophos Fusion, every assessment is informed by live threat intelligence and the collective insight from 625,000+ Sophos-defended organizations, rather than generic benchmarks. Sophos
- [Flashpoint] Ransomware Risk Model – patent awarded. Flashpoint said it’s got a new patent for its ransomware risk model, which scores based on likelihood of vulnerabilities being targeted. The patent covers Flashpoint’s method for rating how likely a vulnerability is to be used in a ransomware event. Vulnerabilities are profiled across more than 60 factors and compared with vulnerabilities historically used in ransomware attacks. Based on their similarity to those patterns, vulnerabilities receive a Low, Medium, High, or Critical Ransomware Risk rating. Flashpoint
- [Progress Software] Better RAG. Progress Software added capabilities to its Agentic RAG platform – a native Microsoft Teams application, support for autonomous multistep retrieval, and a WordPress plugin. The latest Progress Agentic RAG platform release gives organizations two complementary ways to put their knowledge behind AI agents. Content can be ingested in advance through a governed Knowledge Box and kept continuously synchronized for fast, trusted retrieval. In addition, native Model Context Protocol (MCP) support in the platform’s Smart Agent enables agents to access live business systems and web sources at the moment a question is asked, without duplicating data or requiring custom integrations for every system. Together, these capabilities help organizations bring AI into the tools employees use every day, deliver answers that span multiple systems and keep information current, all while preserving governance and avoiding costly rebuilds. Progress Software
- [Exabeam] Agentic SOC vNext. Exabeam extended its Agentic SOC offerings – with autonomous investigations (=faster), guided investigations (=better), and on-premises support (=broader). On the first: Nova AI now works across the platform as a persistent investigator, gathering context, running secondary searches, and retrieving entity profiles as incidents unfold. As measured by our own security operations team, Nova AI triaged an average case in approximately 10 minutes — 30 times faster than the five hours a human analyst would typically require. Related Cases automatically groups connected incidents, so analysts see the broader picture immediately. Exabeam
- [WatchGuard] New cybersecurity goodies for MSPs. WatchGuard announced several major innovations within its Unified Security Platform – for agentic AI-powered security operations, network security (less dumb firewall, more intelligent one), shadow AI discovery, and more. Headlining the announcements is the continued rapid expansion of Rai, WatchGuard’s agentic AI digital workforce purpose-built for MSPs. Unlike AI assistants that simply help analysts work faster, Rai expands operational capacity by taking on assessment, investigation, prioritization, customer reporting, and other complex security tasks across customer environments …. As agentic attacks become more sophisticated, MSPs need more than AI assistance. They need an AI-powered security workforce that helps scale security services without a corresponding increase in operational costs or staffing. WatchGuard