engblogs

A technical case study examining STORM-0817, an Iran-linked malware operation that used AI-assisted debugging, Instagram scraping, and Persian translation to support Android malware development and its command-and-control infrastructure.

A technical case study of CyberAv3ngers, an Iran-linked actor leveraging AI-assisted reconnaissance to exploit ICS/PLCs and weak/default credentials in critical infrastructure.

A technical analysis of the SweetSpecter case study: a China-linked adversary leveraging AI-enabled vulnerability research, LLM-assisted scripting, and spear-phishing to target employees and governments, with MITRE ATT&CK mappings and cross-industry threat intelligence sharing.

Technical analysis of an AI-enabled, multilingual influence operation attributed to Russia, detailing how generated articles, comments, and GAN/DALL·E imagery circulated across UK- and Africa-focused sites and forged information partnerships with local entities.

A technical case study examining how a fake OpenAI-generated error message claiming Russian-troll activity was orchestrated on social media to spread misinformation and trigger mainstream media inquiries.

Case study of Operation A2Z: AI-driven, multilingual influence campaign on social media, its kill-chain orchestration, and disruption through a rapid takedown.

A case study of CyberAv3ngers, an Iran-linked threat actor leveraging AI-assisted reconnaissance and scripting to target industrial control systems and PLCs via default credentials and known vulnerabilities.

Technical case study of Operation Stop News, detailing a Russia-origin influence campaign that used AI-generated multilingual articles, image generation with DALL·E and GANs, and information partnerships with UK entities to influence audiences, followed by platform takedowns.

Technical case study on how Iran-origin actors used AI-generated long-form articles and short social-media comments, posted by fake personas across multiple sites, to influence US and UK elections.

Analysis of OpenAI's takedown of 'A2Z', an AI-powered, multilingual influence operation that automated fake social-media comments on X and Facebook to shape political narratives, revealing how AI-enabled kill chains can be disrupted and prevented.

Case study of a hoax involving a fake OpenAI-generated Russian 'troll' error message, analyzing misattributed JSON content, model usage, and rapid social-media amplification.

An in-depth look at Bet Bot, a scalable operation on X that used AI-generated fake personas and direct messages to spread gambling links, outlining its TTPs and breakout-scale outreach.