Fighting Spy Wars with AI

To its credit, OpenAI published research detailing the explicit use of its platform for espionage and subversion.

Over the last three months, our work against IO actors has disrupted covert influence operations that sought to use AI models for a range of tasks, such as generating short comments and longer articles in a range of languages, making up names and bios for social media accounts, conducting open-source research, debugging simple code, and translating and proofreading texts.

Who was disrupted, and what were they using OpenAI for?

1. Russia, with an operation named Bad Grammar, launched a new operation focused on Telegram and targeting Ukraine, Moldova, the Baltic States, and the United States. The people behind Bad Grammar used our models to debug code for running a Telegram bot and to create short, political comments in Russian and English that were then posted on Telegram.  
2. Russia, with an operation known as Doppelganger, used OpenAI to generate comments in English, French, German, Italian, and Polish that were posted on X and 9GAG; translate and edit articles in English and French that were posted on websites linked to this operation; generate headlines; and convert news articles into Facebook posts. 
3. China, specifically a Chinese network known as Spamouflage, used OpenAI to research public social media activity, generate texts in languages including Chinese, English, Japanese and Korean that were then posted across platforms including X, Medium and Blogspot, and debug code for managing databases and websites, including a previously unreported domain, revealscum[.]com. 
4. Iran, with an operation known as the International Union of Virtual Media (IUVM), used OpenAI to generate and translate long-form articles, headlines, and website tags that were then published on a website linked to this Iranian threat actor, iuvmpress[.]co; 
5. Israel, specifically a commercial company called STOIC, used OpenAI to generate articles and comments that were then posted across multiple platforms, notably Instagram, Facebook, X, and websites associated with this operation.

All the usual malign actors, including Israel, were caught while actively exploring the AI moment in ways entirely contrary to the technology's potential.

As it has been, it will be ... evil do evil until good takes that option away.

Report