OpenAI says it banned ChatGPT accounts after they were used to help place AI-assisted articles under fake journalists’ names in multiple media outlets, CNN reports. The account describes a concrete attempt to use AI for influence, though who commissioned it and whether they were linked to Iran’s government remain unclear.
OpenAI Watch analysis
What happened
CNN says OpenAI reported that about a dozen outlets worldwide published roughly 100 articles credited to seven fake journalists. According to the report, the users prompted ChatGPT in Farsi to revise articles and pitches, then generate social media comments about the coverage. OpenAI said it banned the accounts involved. Read CNN’s report.
The articles included material critical of US policy towards Iran. CNN reports that OpenAI could not identify the particular actor behind the activity and said it appeared consistent with a commercial, for-hire influence campaign. Middle East Monitor’s director told CNN he would investigate the articles OpenAI flagged; CNN requested comment from the LA Progressive, which published one of them.
Why it matters
This was not simply a chatbot producing a misleading answer for one user. The reported operation used AI to prepare material for publication through existing outlets, then to manufacture comments that could make the coverage appear to have an audience. That is a different route to reach: the content can travel with the appearance of ordinary journalism, even when its bylines are fake.
The limits matter too. The reported number of articles and the account of the operation come from OpenAI as relayed by CNN. The reporting does not establish who paid for the campaign, whether a government directed it, or how many readers were persuaded. Those are not small details, and a dateline is not a confession of state involvement.
Our read
AI did not invent propaganda, fake bylines or the old trick of making a message look popular. What it may offer is a cheaper way to draft, adapt and distribute the material. The important question is whether platforms and publishers can spot coordinated deception before the content gains the credibility of a real outlet.
For readers, the practical lesson is to check authorship and sourcing when a striking article starts travelling, rather than treating publication as proof of independence. For platforms, banning accounts addresses the ones identified; it does not, by itself, tell us how much of the operation was disrupted.
What to watch
- Whether OpenAI publishes further detail about how it identified and disrupted the activity.
- Whether the outlets named in the report review or correct the articles in question.
- Whether investigators identify who commissioned the campaign, and whether that changes the account of its reach.
Discussion spark: When AI-assisted influence material passes through real media outlets, should platforms bear the main responsibility for detecting it, or should publishers be expected to verify contributors more rigorously?
Sources and evidence
- Iranian operatives used AI to plant fake stories in multiple US media outlets – CNN (8 October 2026, 20:46 UTC)
OpenAI Watch is independently operated by WittyWires. It is not affiliated with, endorsed by, or operated by OpenAI.