Trump Says Iran Won't Execute Women Protesters—But Iran Says That's Not What Happened

President Donald Trump announced Wednesday that Iran had agreed not to execute eight women protesters after he appealed directly to Iranian leaders. He said the decision came "as a sign of respect" for him.

Iranian officials in Tehran disputed Trump's entire account and said no executions had ever been planned in the first place.

The disagreement is playing out against a wider problem: computer-generated fake posts and videos flooding social media. Both pro-Trump accounts and Iranian state actors are using AI-made content to try to shape what people believe about this diplomatic exchange.

What the Two Sides Are Claiming

Trump initially asked Iranian leaders to release eight women, saying it could be a first step toward broader US-Iran negotiations. He later announced that four would be released right away while four would spend one month in prison.

Iran's court system rejected Trump's claims. Officials said none of the eight women faced execution and called Trump's statements "false news." According to Iranian judicial officials, the women were arrested during anti-government protests in January. If convicted, they would serve prison time at most.

Human rights groups monitoring the situation added another complication: two of the eight women Trump mentioned were already out on bail when he made his initial appeal for their release.

Fake Posts and AI-Generated Videos

The dispute over the women coincided with a bigger problem: an organized campaign of AI-created fake accounts flooding social media. Hundreds of these fake accounts started appearing in recent months, all posting identical messages saying things like "I'm new here and love God, America, and Trump." The messages had awkward grammar that made it clear a computer wrote them.

Trump himself reposted content from at least one of these fake AI accounts, which had a blonde avatar. The post included fake AI-generated photos claiming to show the Iranian women. When people realized the photos were computer-made, the posts became a source of ridicule online.

Iran fired back with its own AI-generated video showing Jesus striking Trump and the former president falling into what looks like hell. Both sides are now using fake computer-made content as a weapon in their messaging fight.

This Has Happened Before

Worth flagging: We have seen this pattern before. When deepfake technology first appeared around 2018-2019, people dismissed it as a curiosity. Within a few years, it became a tool for political manipulation. The AI-generated fake accounts we're seeing now are an upgraded version of those early experiments—but deployed much more widely across social platforms.

The scale is significant. One group of accounts on X (formerly Twitter) that regularly posts AI-generated content has accumulated more than 1 billion views since the Middle East conflict began, according to media monitoring organizations.

A widely shared image showing eight women's faces connected to the Iranian protester story contains fake photos, according to fact-checking tools. When real news reporting gets mixed together with fake computer-made images, it becomes hard for people to tell what's actually true in the moment.

What This Means

Trump also claimed that Iran agreed not to execute "hundreds of political prisoners." He had previously posted about Iran and the protesters that "Help is on the way," which led Iran's exiled Crown Prince Reza Pahlavi to call on the US to follow through on the promise to intervene.

The disagreement about what happened to these women reveals a deeper problem: when countries are separated by distance and distrust, it becomes very hard to verify facts across borders. Iranian officials accused Trump of lying. US officials have not independently confirmed what is actually happening to the prisoners.

Analysis: Both sides have turned to AI-generated content to tell their story. This marks a shift in how countries argue with each other in the digital age. Diplomats still use traditional backchannels to talk, but now they also fight each other using fake posts and videos designed to sway public opinion in real-time. The result is competing stories that may have very little to do with what's actually true.

The Technology Problem

The fake accounts show a sophisticated understanding of how social media works. The fact that hundreds of accounts posted the exact same message suggests someone centrally created the content, while the different fake faces indicate access to advanced AI tools for generating realistic-looking people who don't actually exist.

Fact-checking systems had trouble flagging this fake material as it spread. There's typically a delay between the moment fake content goes online and the moment it gets identified and marked as false. During that window, misinformation can spread widely, especially around breaking news.

The platforms themselves mostly catch problems after the fact through user reports and manual review. They cannot keep pace with automated systems that generate hundreds of fake posts simultaneously. Bad actors using AI have the advantage.

What Comes Next

The dispute over the Iranian women shows how AI-generated content can twist real diplomatic events into something unrecognizable. As tools for making fake videos and images become easier and cheaper to use, more bad actors will be able to create convincing fakes. Meanwhile, the systems to catch and debunk them are falling further behind.

In this author's view, we are entering a period where geopolitics and AI-made content are colliding in consequential ways. When machines can generate thousands of convincing fake images and videos instantly, and platforms spread information faster than anyone can fact-check it, the stage is set for false narratives to shape how people understand international events.

The disagreement over these eight women may matter less than what it reveals: how easily synthetic content can now be weaponized in international conflicts. This sets a worrying precedent for future disputes, where sorting truth from falsehood becomes increasingly difficult.