OpenAI's October 8 report calls the operation "Bogus Bylines." Accounts originating in Iran used seven false journalist personas to pitch English-language articles, mostly about the US-Iran conflict, to small and medium-sized publications. About 100 of the pieces ran between July 2025 and October 2026. The outlet count is unclear: the Post says at least 20 news outlets, and OpenAI's report says roughly a dozen. The operators had ChatGPT edit the articles, fit them to each publication's submission rules and write pitch emails to editors. OpenAI said it looked like a commercial, for-hire influence campaign. The operators went through VPNs, even though OpenAI already limited access from Iran. Some pieces reached outlets through syndication networks like PeaceVoice. In 2024, Iranian ChatGPT activity did not appear to get meaningful audience engagement. This time, OpenAI rated the article placements Category 4 on a six-point scale. For AI builders, this shows that geoblocking alone does not stop misuse. Catching it took behavioral analysis across accounts. Model providers should expect platform-abuse reporting to become a standard part of what they offer.