engblogs

A technical case study of a Russia-origin influence operation that used AI-generated multilingual content and coordinated information partnerships to reach audiences across English-, French-, and Russian-language platforms.

A technical case study detailing Iran-origin influence operations that used AI to generate English and Spanish election content and coordinated long-form articles with social media comments across X, Instagram, and multiple websites, and the disruption efforts by OpenAI.

Case study of OpenAI banning Rwanda-origin accounts that used AI to flood X with high-volume, multilingual, hashtag-driven partisan comments ahead of elections.

Case study of OpenAI's A2Z multilingual influence operation: an AI-enabled campaign of fake social-media personas and multilingual comments on X and Facebook to sway elections and Azerbaijan-related narratives, revealing AI-driven kill-chain vulnerabilities and disruption tactics.

Bet Bot: A technical deep-dive into how AI-generated personas were used to seed conversations and direct messages on X to broadcast gambling links, revealing abuse vectors, TTPs, and detection challenges.

Analysis of Operation STORM-2035: Iran-origin influence activity leveraging AI-generated long-form articles and short social posts to sway US politics, detailing dual workstreams, cross-platform posting, and detection challenges.

Technical analysis of the Tort Report case study on AI-generated abusive reporting targeting Vietnamese media, examining detection, account bans, and the minimal impact on Facebook and YouTube moderation.

Case study of an AI-generated partisan commenting network that produced high-volume, hashtag-driven content ahead of Rwanda's elections, and how OpenAI detected and banned Rwanda-origin accounts.

Case study of OpenAI banning API-based accounts that powered the 'Bet Bot' gambling-spam network on X, using AI-generated fake personas to post public comments and send direct messages with gambling links.

Tech case study on an API-driven operation that used AI-generated comments to attack the Anti-Corruption Foundation (FBK) and Navalny on X, via fake accounts and multilingual replies, with a modest impact assessment.

A technical case study on how AI-enabled accounts from Rwanda generated high-volume, hashtag-driven partisan comments on X ahead of elections, how OpenAI detected and banned the activity, and what this reveals about scalable content generation and comment-spamming networks.

Technical analysis of an AI-assisted abusive reporting operation ('Tort Report') targeting Vietnamese independent media on Facebook and YouTube, detailing its methodology, signals, and observed impact.