🛡️ InfoSec Blue Team Briefing

Tuesday, November 17, 2026

🎧 Audio Briefing

Download MP3

Here's what the BlueTeamSec community was sharing on infosec.pub for Tuesday the 17th of November 2026. We've got an AI detection engineering paper, a podcast about some very unglamorous attacks, and a long look back at Gootloader.

Averlon offers commentary on a paper called Unconstrained CTI-REALM with Open-Weights, which finds that open-weight models match the proprietary frontier ones at automated detection engineering once you stop capping their tokens and messages. The practical advice is to route work across specialised models, check whether your SOAR agents are failing because of the constraints you gave them, and judge generated queries and Sigma rules by whether they actually run. Worth a look if you're building detection pipelines, and it's another entry in the 'AI changes everything' file, though this one at least comes with evidence.

The Ahead of the Threat podcast, in its third season, has FBI, Dutch and UK cyber authorities discussing a fourteen-year-old unauthenticated remote code execution flaw in PHP, found independently by several researchers using frontier AI models. An attacker apparently tried to exploit it shortly before public disclosure. They also cover attacks on web-facing refrigeration controllers at hundreds of French supermarkets, where default credentials let attackers trigger defrost modes that spoiled food and medical supplies, and similar tactics have turned up at US water utilities. Default passwords on internet-facing kit in 2026 is a sentence that should be extinct by now. We touched on exposed AI pentest consoles in the SOCRadar piece yesterday, if you want a neighbouring theme.

The gootloadersites blog has a retrospective on five years of chasing Gootloader, the network that used SEO poisoning and hijacked WordPress sites as proxies to backend servers delivering zipped JavaScript. It acted as an initial access broker for ransomware crews and kept hardening its backdoors against researchers. The write-up covers the reverse engineering and intelligence sharing behind its disruption in Operation Endgame, which it dates to the 12th of November. Worth reading in full if you like your threat research with a long memory.

That's the briefing for today. Any attribution of activity to a country, group or organisation was the original authors' claim rather than ours, and the analysis behind these summaries was automated. Thanks for listening.

📰 Articles Covered