OpenAI Blocks Iranian and Russian AI Influence Campaigns
OpenAI said it disrupted two long-running influence operations, one tied to Iran and one to Russia, that used its AI models to spread anti-American and anti-Israeli messaging. The Iranian campaign created fake identities for at least seven Western journalists and successfully placed false articles in smaller and midsize news outlets. OpenAI rated the Russian operation higher than the Iranian one on its scale for potential reach, noting both reached mainstream media.
OpenAI traced the Iranian effort, called “Bogus Bylines,” back roughly two and a half years. Operatives invented at least seven Western journalist identities and used them to place fabricated articles in smaller and midsize outlets. The company identified AI involvement because chatbot prompts were submitted in Persian.
The Russian campaign received a 5 on OpenAI’s six-level reach scale, while the Iranian one received a 4; both reportedly reached mainstream outlets. Since July 2025, the seven fake journalists produced 100 fabricated pieces. Some outlets, including Daily Kos and Middle East Monitor, removed content after being contacted.
This disclosure may heighten concerns among journalists, newsrooms, and audiences about verifying bylines and sourcing. Smaller outlets with limited fact-checking resources could be especially exposed, while readers may become more skeptical of opinion pieces and conflict coverage. If AI-assisted influence campaigns continue, platforms and publishers may face pressure to strengthen identity checks and disclosure practices. The broader public could encounter more polished disinformation, making media literacy and trust-building efforts more important.