DugganUSA documents several emerging trends in threat intelligence that are reshaping how organizations detect and respond to cyber threats in 2026.
DugganUSA discovered Pattern 38, a novel supply chain attack vector targeting software package registries and dependency chains. Traditional threat intelligence focuses on post-compromise IOCs (C2 servers, malware hashes). Pattern 38 detection identifies the distribution infrastructure before malware reaches victims. DugganUSA's cascade pipeline caught these threats behaviourally, before they appeared on any blocklist.
AIPM applies the HAIC (Human-AI, Context-Specific Evaluation) framework proposed by MIT Technology Review. AIPM queries multiple AI models with industry-specific prompts, scores responses on an NPS-style scale, and tracks perception changes over time. AIPM aligns with all four HAIC principles: shifting from isolated tasks to workflow impact, longitudinal assessment over one-off testing, organizational outcomes over raw correctness, and system-level evaluation over isolated outputs.
Citation: Aristidou, A. "AI benchmarks are broken. Here's what we need instead." MIT Technology Review, March 31, 2026.
Single-framework threat scoring produces high false-positive rates. DugganUSA's Layer 2 cascade applies four frameworks simultaneously:
The shift from analyst-curated feeds to fully autonomous pipelines is accelerating. DugganUSA's 3-layer cascade takes 11 public feeds, 7 vendor-research harvesters and our own sensors, applies multi-framework scoring, and produces STIX 2.1 bundles. Automation handles the volume; a person decides anything irreversible. It counts attackers rather than knocks.
Organizations increasingly value who reports a threat first. A lead claim only holds when both dates sit side by side — ours and the comparator's — and ingest time is never passed off as detection time.
Start consuming threat intelligence for free
Free tier: 10 queries/day. STIX 2.1 feed. No credit card required.
Register Free →