The AI Sob Stories Flooding X Are a Business. Here Is Who Is Getting Paid.

Key points

• X's Creative Revenue Sharing programme pays creators whose posts rack up millions of impressions, and some are using AI to manufacture fake personal stories purely for engagement.
• One creator in Nigeria told Wired AI he earns between $500 and $700 every two weeks posting AI-written melodramas; some others report earning thousands of dollars per payment cycle.
• In May 2025, X sued eight Vietnamese creators and 25 unnamed individuals, alleging they fraudulently claimed programme money using AI-generated content boosted by bot-driven comments.
• On 16 July 2025, X said it had removed millions of posts that copied already-published content from other users.
• Researchers say the stories follow a rigid structure: wronged underdog, dramatic twist, villain publicly humiliated, and the format is trivially easy to prompt an AI to produce.

A nine-year-old walks into a federal courtroom carrying a blue backpack and tells the judge he knows who framed his mother. Hundreds of thousands of people click. Nobody in that story exists.

That post, and thousands like it, come from creators who feed a short prompt into an AI writing tool, such as ChatGPT or Grok, and let the software do the rest. The resulting story gets posted to X as though it were a real person's lived experience. Views pile up. X pays out.

How does the payment actually work?

X's Creative Revenue Sharing programme pays creators whose posts drive engagement on the platform. To qualify, a creator must subscribe to X Premium, hold at least 500 verified followers, and have accumulated a minimum of five million impressions in the previous three months. X says the aim is to reward "authentic, high-quality content."

Paying by engagement, though, rewards whatever gets people to click and comment, true or not.

The creator behind the courtroom story goes by the handle BIGBEN. He told Wired AI, the outlet that first reported this story, that he is a 21-year-old stock trader living in Nigeria. He uses Grok or ChatGPT to write the posts. A recent story he published, narrated by a second wife mistreated by her in-laws, reached 1.5 million views in two days. His typical fortnightly paycheque from X: $500 to $700.

Why are these stories so hard to stop?

They are hard to stop because they are hard to prove harmful. The stories are fiction dressed as fact, but many creators describe themselves as storytellers in their profiles. No single law is obviously broken.

Rob Cover, a professor at RMIT's Digital Ethnography Research Centre in Australia who studies AI-generated content, says the format is as rigid as a sonnet. First, an innocent person is wrongly fired, evicted, or accused of a crime. Then a twist, often a child with a recording device or a sealed document, exposes the villain. Finally, the villain is publicly shamed. Repeat.

Jenna Russell, a PhD candidate at the University of Maryland researching AI writing, puts it simply: "These spammers have optimised for the least effort and the most engagement, which means using easy, well-known tropes and prompting AI to write something that fits the bill."

X has taken some action. In May 2025, it sued eight Vietnamese creators alleging fraudulent use of the programme. In April it cut payments to "aggregators" who mostly reposted others' news. The company, which is now part of xAI, itself recently merged with SpaceX, did not respond to questions about AI-generated fiction on the platform.

What does this mean for ordinary readers?

These stories feel real because they are designed to. They tap a genuine human need: the wish to see fairness win. That jolt of satisfaction when the villain is humiliated is real, even when every detail is invented.

For most readers the practical harm is modest. You waste a few minutes; the algorithm learns you enjoy this genre and serves you more. The bigger concern is a slow blurring of the line between a real person's story and a machine's product, published purely to extract a payment.

If a post makes you feel a sharp, immediate surge of outrage or vindication, it is worth pausing. That emotional shape is exactly what the format is engineered to produce.

Common questions

Is any of this illegal?

Not obviously, at present. Posting fiction is not fraud on its own. X's lawsuit against the Vietnamese creators rests on allegations of bot-boosted engagement rather than the AI writing itself.

Can X's own AI tools detect this content?

AI detectors, software that tries to identify machine-written text, remain unreliable. Researchers say creators can tweak prompts to evade them, and false positives would catch genuine writers too.

Should I change how I use X because of this?

No dramatic action needed. Being sceptical of vivid first-person stories that end with perfect justice is a reasonable habit on any platform, not just X.