Spanish opposition deepfake of PM's wife draws outrage before vote

Spain's main opposition party, the conservative People's Party (PP), has drawn sharp criticism over a political advertisement that used a deepfake video of the wife of Prime Minister Pedro Sánchez. According to The Guardian, the video was launched this week across the party's social media channels and has provoked outrage in a country already deeply divided ahead of its general elections next month.

The advertisement used artificial intelligence to alter the likenesses of several people connected to Sánchez's Socialist party. In the clip, those figures appear to confess to misconduct, including soliciting prostitutes and accepting kickbacks. The Guardian reports that the video has set off a broader conversation about how AI-generated media should be used in political campaigns.

The report does not describe the content as verified or the allegations as true; the footage itself is synthetic, which is the central point of the criticism. Based on the outlet's reporting, the controversy has centered on the decision to present fabricated confessions by real people in a paid political message.

The timing adds to the tension. The election campaign is entering its final stretch, and the use of realistic synthetic video in an attack advertisement raises questions about how voters will tell authentic footage from manipulated material. The Guardian's account makes clear that the reaction has been strong, though the piece does not detail official responses in the material available.

Background

Deepfakes are media in which software replaces or synthesizes a person's face, voice or body so that they appear to say or do something they never did. The technology has improved quickly over the past several years, making convincing fabrications far easier and cheaper to produce than they were when the first widely circulated examples appeared. Early concerns focused on non-consensual pornography, but political misuse has become an increasingly common worry as elections in many countries have taken place in the shadow of synthetic media.

Political advertising has long relied on exaggeration, selective framing and attack messaging. Most democracies allow sharp criticism of opponents, and satire has a long and protected history. What makes synthetic video different is the sense of direct evidence it carries. A written claim can be disputed on its face, but a moving image of a person apparently confessing feels like documentation, even when it is fabricated. Researchers who study misinformation have repeatedly noted that corrections often arrive too late or reach fewer people than the original clip.

Spain has its own regulatory context. The country has debated rules on electoral communication and online political content, and the European Union has been developing a broader framework for artificial intelligence, including transparency duties for certain synthetic content. Rules differ by country, and enforcement is often slower than the pace of new tools. Where campaign advertising is involved, the line between protected political speech, defamation and election-integrity concerns can be contested, and different legal systems weigh these differently.

The broader pattern matters too. Parties across the political spectrum have experimented with generative tools for posters, videos and messaging. Some use them openly and label the output as synthetic, while others blur the distinction. A labeled illustration of a rival is one thing; a realistic simulation of a real person confessing to crimes is another, because it borrows the credibility of real footage to deliver an accusation that never happened.

Why it matters

This episode goes beyond one party and one campaign. It tests whether existing norms about political messaging can hold when fabricated video becomes easy to make and fast to share. If a campaign can attach invented confessions to real officials or their relatives and face only public anger, the incentive to repeat the tactic grows. Each use also lowers the baseline trust that voters place in any video, including authentic footage that later needs to be defended.

The targets of such material are not only the politicians depicted. People linked to a party, including family members and staff, can become subjects of harassment and reputational harm without having any role in the campaign. The people shown in the video are real individuals whose likenesses were used to deliver accusations, which raises consent and dignity questions that political rhetoric rarely faces.

For democratic debate, the risk is less about any single viewer believing a specific fabrication and more about a general fog of doubt. When people cannot tell what is real, they may dismiss legitimate evidence of wrongdoing as "probably AI," which can weaken accountability. Election bodies, courts, platforms and parties all have a stake in preventing that outcome, and each faces pressure to respond quickly and credibly.

The case also shows how social media distribution shapes the stakes. A clip posted across several channels at once reaches audiences that a traditional advertisement would not, and removal or labeling after the fact may not undo the initial impression. How platforms and campaigns handle disclosure, and whether voters receive clear signals about synthetic content, will influence how damaging similar episodes become.

What to watch

Watch for any formal complaints, legal challenges or regulatory responses in Spain, and whether electoral authorities or courts address the advertisement directly. Note whether the People's Party removes the video, adds a disclosure label, or defends it as satire or commentary, since each response sets a different precedent.

Pay attention to how platforms treat the clip under their policies on synthetic political content, and whether other parties or campaigns adopt similar tactics before the vote. Reporting that tracks the spread of the video and the responses it draws will help clarify how far the controversy reaches beyond the initial outrage.