Spain's PP faces a criminal complaint over its AI deepfake election video: what's alleged

Spain's PP AI deepfake video has become a legal fight in the country's snap election campaign. On Wednesday, October 21, 2026, the conservative People's Party posted a video in which AI-generated likenesses of six Socialist-linked figures appear to confess to wrongdoing. Two days later, one of them, former Socialist party member Leire Díez, filed a criminal complaint against the party at Spain's National Court. The PP says the video is satire and is labeled as AI. Under the EU AI Act, deepfakes have had to be disclosed since August 2, 2026, as our AI Act explainer sets out.

This story is based on reporting by Vozpópuli, The Objective, Ara, elDiario.es, El Español and Telecinco, the Spanish National Statistics Institute's election page, and the text of the AI Act's Article 50, checked October 26, 2026. Spanish quotes are 9to5AI's translations. The complaint's allegations haven't been tested in court, and nothing here is legal advice. The story was first flagged by an aggregator; every detail here comes from the outlets above.

The short version

• The video: posted October 21, 2026, on the PP's official accounts. AI versions of Begoña Gómez, José Luis Rodríguez Zapatero, José Luis Ábalos, Leire Díez, Montse Mínguez and Santos Cerdán each "admit" to an alleged misdeed.
• The complaint: Leire Díez's querella, filed with National Court judge Santiago Pedraz on October 23, alleges slander with publicity, an offense against moral integrity and digital identity theft.
• The PSOE: the Socialist party said on October 22 it was studying legal action. No filing had been reported as of October 26.
• The PP's defense: it's satire, and the post says it's AI-made.
• The election: a general election on November 29, 2026.

What did the PP's AI video show?

The video parodies a PSOE post in which Prime Minister Pedro Sánchez's party called itself "zurdos y orgullosos," "lefties and proud," ahead of Argentine President Javier Milei's visit to Madrid. In the PP's version, each AI figure is made to give a reason for being a "zurdo," according to Telecinco's October 22 account. These are the video's insinuations, not established facts:

• Begoña Gómez, Sánchez's wife, because she got a university post thanks to her husband;
• José Luis Ábalos, a former minister, because he pays for prostitution and takes illegal commissions;
• José Luis Rodríguez Zapatero, a former prime minister, because he "hides jewels";
• Leire Díez, because she hired thugs to get rid of a judge, a reference to Judge Beatriz Biedma;
• Montse Mínguez, a former PSOE spokesperson, over her partner's property holdings;
• Santos Cerdán, a former PSOE organization secretary, over bribes.

The closing slogan becomes "lerdos y orgullosos," "dimwits and proud." Telecinco counted more than 1 million views on X, more than 500,000 on Instagram and more than 200,000 on TikTok by October 22. The words the video puts in the figures' mouths are AI-generated, not things they said.

What does Leire Díez's complaint allege?

Díez, represented by lawyer Eva Bejarano, filed a querella, a criminal complaint brought by a private party, against the PP as a party, with no individuals named, according to Ara and The Objective. Vozpópuli reported that it went to Judge Santiago Pedraz of the Audiencia Nacional, who is investigating a case involving Díez in which the PP is a popular accuser. It alleges:

• calumnias con publicidad, slander with publicity;
• an offense against moral integrity;
• usurpation of identity in the digital sphere.

The complaint says the video was made "expressly and intentionally to humiliate her" and calls its content "flatly false," Ara and Vozpópuli reported. It asks for the video's provisional removal from all PP accounts. Reports of how many views it cites differ: Vozpópuli put the defense's estimate at more than 2 million, while The Objective reported more than 21 million. No outlet had reported whether the court admitted the complaint as of October 26.

What do the PP and the PSOE say?

PP national spokesperson Borja Sémper said that if the video were meant "to make people believe" the figures were real, "we would have a problem," but that viewers understand it as satire about Socialist hypocrisy, El Español and elDiario.es reported. "What worries me is that there are insults, not AI," he said. The PP's spokesperson in Congress, Ester Muñoz, said the video is plainly AI and that it "tells the truth," and criticized the PSOE for wanting to "judicialize politics."

elDiario.es reported that the post carries a label reading "contenido hecho con IA," "content made with AI." The PSOE's organization secretary, Rebeca Torró, said the PP "use technology" because "they have nothing but insults," and party sources said legal action was under study, elDiario.es reported.

What does the EU AI Act require for deepfakes?

Article 50(4) of the AI Act says deployers of an AI system that generates or manipulates image, audio or video content "constituting a deep fake" must "disclose that the content has been artificially generated or manipulated." The duty has applied since August 2, 2026.

There's a carve-out that fits the PP's argument. Where the content is part of "an evidently artistic, creative, satirical, fictional or analogous work," the duty is limited to disclosing it "in an appropriate manner that does not hamper the display or enjoyment of the work." Article 50(5) adds that the disclosure must be "clear and distinguishable" and come no later than "the first interaction or exposure."

So the AI Act question is narrow: whether the label was adequate, and whether the video counts as evidently satirical. Díez's complaint rests on Spanish criminal law, and an AI label doesn't settle a slander or identity claim. No Spanish authority had said whether the video meets Article 50, and no complaint to Spain's Junta Electoral Central had been reported as of October 26.

How does this compare with the US?

US rules are patchier. There's no federal AI disclosure rule for political ads ahead of the November 3, 2026 midterms; states set their own, as our guide to AI political ad labels explains. In Brazil, the electoral court has gone further, ordering Meta to make its AI tool treat candidates equally, as our Brazil story reports.

What's confirmed, reported and uncertain

• Confirmed: the November 29, 2026 election date and Royal Decree 806/2026 (INE); the Article 50 text.
• Reported: the video's content and view counts (Telecinco, El Español); the querella, its court, charges and requests (Vozpópuli, The Objective, Ara); the AI label (elDiario.es); the PP and PSOE statements (El Español, elDiario.es).
• Uncertain: whether the court will admit the complaint or order removal; whether the PSOE will file; how big and visible the AI label is; and how Spanish authorities would apply Article 50's satire carve-out.