All news with #detection engineering tag

🔍 Traditional SIEM logic — fixed rules that match event A followed by event B — is increasingly insufficient against modern, sophisticated threats that use legitimate tools and supply-chain vectors. Kaspersky describes a shift to continuously updated correlation content informed by its MDR service and threat research. In 2025 the team delivered dozens of updates and hundreds of new or refined rules, and now maintains over 850 rules mapped to MITRE ATT&CK. Integration with Kaspersky EDR and expanded telemetry helps detect multi-stage attack chains and reduce false positives.

🔍 In a sweeping analysis of 25 million enterprise security alerts, researchers found that nearly 1% of confirmed incidents began as low‑severity or informational alerts, rising to about 2% on endpoints. The dataset included 10 million monitored endpoints, 82,000 forensic endpoint investigations with live memory scans, and 180 million files analyzed. The report shows EDR remediation frequently reports systems as 'mitigated' even when memory forensics reveal active malware, and it documents evolving phishing and cloud persistence tactics that evade legacy triage models.

🤖 Enterprises migrating between SIEM platforms face repetitive, error-prone rule rewrites because vendors like Splunk, Microsoft Sentinel, IBM QRadar, and Google Chronicle use distinct query languages and data models. Researchers from the National University of Singapore propose ARuleCon, an AI-assisted framework that translates rules while preserving detection intent. In tests on nearly 1,500 conversions it improved accuracy about 10–15% over baseline LLM approaches. Practitioners caution that deterministic engineering, robust validation, and human oversight remain essential to avoid semantic drift and operational risk.

⏱️ Unit 42 data and a conversation with Wendi Whitmore warn that attackers can exfiltrate data in as little as 39 seconds, forcing a shift from prevention to rapid detection and containment. Whitmore argues manual workflows cannot match adversary tempo and calls for AI-driven detection paired with unified visibility across endpoints, cloud and AI systems. Visibility, not complexity, enables containment before escalation.

🔍 The 2026 Unit 42 Global Incident Response Report warns that adversaries are moving to exfiltration four times faster than in 2025 and are exploiting gaps created by an over-reliance on endpoint telemetry. Unit 42 found critical evidence present in logs for 75% of incidents, yet siloed systems and inaccessible telemetry prevented timely detection and response. The authors recommend a single-pane-of-glass, AI-driven SOC that centralizes logs and uses tools like Cortex XSIAM for alert stitching, ML-based scoring and unified investigations to reduce alert fatigue and close multi-surface blind spots.

🔒 As AI tools such as Claude Mythos and Project Glasswing compress vulnerability discovery from weeks to minutes, the traditional patch window is effectively gone. The piece urges organizations to adopt an assume-breach posture that prioritizes rapid detection, automated attack reconstruction, and immediate containment. Network Detection and Response (NDR) platforms — highlighted via Corelight — are presented as practical instruments to visualize, measure, and reduce mean-time-to-contain.

🔍 The UK National Cyber Security Centre (NCSC) cautions that many common SOC metrics are misleading and can actively harm security operations if used or reported externally. CTO Dave Chismon argues that only time to detect/time to respond (TTD/TTR) reliably demonstrates SOC effectiveness, while metrics such as ticket counts, closure times, rule counts or raw log volume create perverse incentives. He recommends red and purple team exercises to assess TTD/TTR, and suggests internal, non-public metrics — hypothesis-led hunting, strict false-positive thresholds, log coverage, tooling expertise and analyst engagement — to monitor week-by-week health without driving the wrong behaviours.

🔍 AI-driven detection reduces alert noise and accelerates incident identification by building behavioral baselines across users, endpoints, identities, and cloud workloads. Platforms that combine behavioral models, cross-telemetry correlation, and automated triage suppress low-value alerts, enrich context, and prioritize what matters for lean security teams. Paired with managed detection and response, integrated automation shortens dwell time, limits lateral movement, and reduces operational impact when prevention fails.

🔐 Google announced a shift to agent-centric security at Google Cloud Next '26, positioning AI agents to help SOC teams respond to the potential surge of vulnerabilities tied to Anthropic's Mythos. It introduced three new agents in Google Security Operations — a threat hunting agent, a detection engineering agent and a third-party context agent — and said its existing triage agent has processed over five million alerts, reducing analysis from about 30 minutes to roughly a minute with Gemini. Additional moves include expanded Wiz integrations, an AI-BOM to inventory AI components, agentic automation features, Model Armor protections, Agent Identity and Agent Gateway controls, and modern IAM simplifications to streamline permissions.

🔍 Mature SOCs compress MTTR by embedding threat intelligence directly into analyst workflows rather than relying on separate feeds, reports, or manual lookups. The contributed piece from ANY.RUN outlines five operational areas—detection, triage, investigation, response, and threat hunting—where integrated TI Feeds, TI Lookup, and Threat Reports remove handoffs. By surfacing behavioral context and enabling SIEM/SOAR automation, teams detect earlier, decide faster, and contain threats with minimal delay.

🔍 Artificial intelligence is transforming how security teams detect and hunt threats by processing vast telemetry at scale, correlating noisy signals, and surfacing behavioral anomalies faster than traditional tools. Organizations report efficiency gains—often 40–50% on lower-tier SOC tasks—as AI automates alert triage, log review, documentation, and evidence collection. Vendors say AI reduces alert fatigue by clustering and prioritizing incidents, but experts stress a human-in-the-loop approach and strong governance to avoid amplifying weak security practices.

🔒 Attackers are compressing the time from initial access to lateral movement by using AI, automation and refined TTPs, forcing defenders to adopt prevention-first strategies. The article highlights that average breakout time is about 30 minutes and that exfiltration can sometimes occur in minutes, with extreme cases measured in under ten minutes. It recommends AI-powered XDR/MDR, unified visibility across endpoint, network and cloud, and stronger identity-centric controls to speed detection and response. Automated containment—session termination, host isolation and password reset—should be orchestrated with SIEM and SOAR to reduce dwell time.

🔍 In 2025 Japan reported 134 ransomware incidents—a 17.5% increase from 2024—with Qilin responsible for 22 cases (16.4%). Talos highlights Qilin’s growing automation, credential‑based access, and use of an EDR‑killer that targets 300+ drivers and employs locale-based geo‑fencing. The blog focuses on detecting activity during the pre‑ransomware phase (average six days to execution) and shares Sigma/YARA rules plus correlation guidance to reduce false positives.

🔒 CrowdStrike outlines detection for CVE-2026-20929, a Kerberos relay vulnerability exploited via DNS CNAME abuse that can enroll certificates from Active Directory Certificate Services (AD CS). Their correlation-based detection flags anomalous certificate-based authentications coincident with unusual AD CS Kerberos service access within a short time window. Customers can enable the provided CRT rule in Falcon Next‑Gen SIEM to activate alerts and support hunting.

🔍 Modern attackers use AI to imitate trusted users, tools, and services, making many incidents malware-free and harder to detect. The article compares these tactics to art forger Elmyr de Hory and outlines threats such as agentic AI, supply-chain impostors, cloaked tunnels, rogue infrastructure, and sophisticated phishing. Network Detection and Response (NDR), including Corelight’s Open NDR Platform, is highlighted as essential for spotting behavioral anomalies, protocol inconsistencies, and contextual metadata to expose impostors early.

🔍 Microsoft introduces CTI-REALM, an open-source benchmark that evaluates AI agents on end-to-end detection engineering by turning real-world cyber threat intelligence into validated detections. The benchmark places agents in realistic, tool-rich environments where they must read CTI reports, explore telemetry, iterate on KQL queries, and produce Sigma rules and KQL-based logic scored against ground truth across Linux, AKS, and Azure. CTI-REALM's checkpoint-based scoring surfaces whether failures arise from CTI comprehension, technique mapping, data-source selection, or query construction, helping teams decide where human oversight and guardrails are required.

🔍 Cisco Talos introduces DispatchLogger, an open-source DLL that transparently instruments late-bound COM (IDispatch) interactions to enhance malware analysis visibility. The tool hooks COM instantiation APIs and returns proxy objects that forward calls while logging method names, parameters, return values, and object relationships. It supports recursive wrapping, enumerator proxies, and moniker handling to reveal high-level automation events often missed by low-level API tracing. Deployment requires injecting the DLL into target processes and preserves COM lifetime and threading semantics.

🔐 Modern phishing increasingly hides behind legitimate infrastructure and encrypted HTTPS, making static checks insufficient. The piece recommends a three-part investigation model — safe interaction, automation, and in-sandbox SSL decryption — so SOCs can observe full attack flows, extract actionable IOCs, and reach evidence-based verdicts quickly. This approach reduces analyst load and helps detect identity-driven compromise earlier.

🔍 Cloudflare's Log Explorer centralizes 14 new datasets to give analysts correlated, edge-to-core telemetry for investigating multi-vector attacks. By combining HTTP requests, Firewall, Zero Trust access, IDS, DNS and gateway logs, teams can rapidly reconstruct reconnaissance, exploitation, and exfiltration chains. The platform reduces detection time and supports schema-driven ingestion for future data sources. It also improves ingestion latency and enables concurrent queries for faster, correlated forensics.

🔍 Cloudflare’s redesigned Security Overview dashboard helps security teams turn overwhelming telemetry into prioritized, actionable remediation. The interface introduces Security Action Items — ranked by Critical, Moderate, and Low — alongside a Detection Tools module that indicates whether protections are actively enforcing or left in "Log Only" mode. Suspicious Activity cards deep-link into Security Analytics to preserve filters and speed triage.