An Iraqi MP's face was put on a dancer's body, a fact-check proved it, and the video stayed up

What happened

Aliya Nasif was campaigning for re-election to the Iraqi parliament in October 2025 when someone used AI face-substitution tools to put her face onto the body of a Ukrainian dancer, in a video originally posted to TikTok in 2023. The clip was framed as a pornographic scandal and posted to X.

Checker News, an Iraqi fact-checking platform, documented over 150,000 views on a single account, and SMEX reports that other pages amplified it across platforms. Nasif, a sitting member of parliament, publicly denied it. The fact-check established the fabrication. SMEX’s next sentence reads: “The content remained.”

Everything worked except the outcome

Read the sequence again. The fabrication was caught during the campaign, quickly, by a local organisation, in the right language, and published. This is the scenario that most guidance about synthetic media is written to produce, and it produced nothing.

That inverts the usual planning assumption. Verification is normally treated as the hard part, with removal following from it once the facts are established. Here verification was the part that succeeded.

SMEX’s explanation points at the platforms rather than the fact-checkers. By 2024 the dominant platform response to AI-generated content had converged on labelling rather than removal, a posture that puts the burden on metadata and creator disclosure. And the enforcement systems that apply the written rules do not function equally across languages and regions: the rules against non-consensual intimate imagery and synthetic abuse exist at Meta, X, TikTok and YouTube, and Arabic sits at the periphery of where moderation money goes.

SMEX also records a difference in what gets targeted, citing researchers who have documented it globally: when AI is used against male candidates the target is their political credibility, and when it is used against women it is their bodies.

What this changes about a response

Plan the response for a world where the content stays up, because in the one documented Iraqi case we have, it did. A takedown is a hoped-for outcome, not a step in the plan.

Then do the fact-check anyway, and fast, but be clear with the person about what it buys. Its value here is not removal. It is a dated, public, third-party record that the person can point at afterwards, for years, when the clip resurfaces in front of an employer, a family, a court or an election commission. Nasif still has that. She does not have a takedown.

Decide in advance who speaks. Two separate things happened here: Nasif denied the video herself, and a fact-checking organisation established the fabrication independently. They do different work. One is her testimony. The other is a record a stranger can cite without having to take her word for anything.

What is not in this account

Who made the video. Who posted it first. Total reach across the other pages. Any platform response at all. Whether the clip is still online now. And the fact-check itself, which we could not open.