AI-generated images in paid X ads show economist Gonzalo Bernardos in invented arrest and corruption scenes to promote an unauthorised investment platform, and Bernardos says a video made with AI advertises a supposed investment; Bernardos has warned on X since July 2026
Maldita.es reported on 1 October 2026 that it found at least 20 paid ads on X, circulated on 24 and 25 September 2026, that used AI-generated images to show the economist Gonzalo Bernardos in invented scenes, among them police searching the economist's home. The ads were seen more than 400,000 times and, depending on the device, led to pages imitating El Mundo that said Bernardos had been arrested and promoted Norden Core, a platform not authorised by the CNMV. Marca reported that Bernardos warned on X on 23 July 2026 that individuals were using the economist's image in fake ads made with AI or other tools. El Caso reported on 4 October 2026 that Bernardos said on X that many people were asking about a video advertising a supposed investment, that the video was made with AI and that Bernardos had never made such promotions. No source reports a person who lost money or a complaint to the police.
- AI system
- Unidentified image and video tool
- Occurred
- 2026
- Reported
- 24 July 2026
- Event location
- Unknown
- What the AI did
- Depicted or impersonated the person
- Reported harm
- Reputational Harm
- Whose AI use
- Someone else’s AI use
- Setting
- Finance · Work
- Evidence
- AI involvement reported · Causal attribution supported · 3 sources, 2 underlying accounts
- 6 claims: 6 reported. 5 open questions
- People reported harmed
- At least 1 person
AI system as recorded: AI-generated images in paid ads on X (Maldita.es says OpenAI's tool for identifying content generated with its own tools detected the SynthID digital watermark in all of them) and a video that Bernardos says was created with AI; the generating tools and the operators are not identified
What Happened
Maldita.es (1 October 2026) analysed at least 20 paid ads that circulated on X on 24 and 25 September 2026. The ads carried AI-generated images that appear to involve Bernardos in corruption scenes: police searching the economist's home, the economist waiting in a park with a bag full of banknotes, and the economist burning papers in the street with a television presenter. Maldita.es says it checked that all the images were generated with AI and that OpenAI's tool for identifying content generated with its own tools detected the SynthID digital watermark in all of them. The ads were seen more than 400,000 times in total. Their link appeared to point to Amazon but, depending on the IP address or location, led to pages imitating El Mundo which said Bernardos had been arrested for promoting an investment platform and which promoted Norden Core, a platform not authorised by the CNMV. Maldita.es located the advertising accounts in the United Kingdom, Germany, Brazil, Lithuania and Turkey, and reports that X removed most ads within hours while new ones appeared with the same images.
Marca (24 July 2026) reported an earlier warning: on the night of 23 July 2026 Bernardos wrote on X that individuals were using the economist's image to advertise investments and financial products, that all were fake ads using AI or other tools, and asked people to ignore them. El Caso (4 October 2026) reported that Bernardos said on X that many people were contacting the economist about a video in which the economist's image announces a supposed financial investment, that the video was created with AI and that Bernardos had never made such promotions.
Reported harm
AI-made images in paid ads showed Bernardos in invented arrest and corruption scenes to promote an unauthorised investment platform (Maldita.es), and Bernardos says an AI-made video advertises an investment the economist never promoted and that many people ask about it (Bernardos on X, as paraphrased by El Caso; Marca quotes the earlier warning of 23 July 2026 about fake ads).
Outcome
UnknownBernardos warned followers on X on 23 July 2026 (Marca) and again in a post that El Caso reported on 4 October 2026. Maldita.es reports that X removed most of the ads within hours for breaking its rules, while two had not been removed as of 1 October 2026. No police complaint, lawsuit or defrauded person is reported.
What remains unknown
- Who paid for the ads and which tools made the images and the video.
- Whether anyone lost money through the promoted platforms.
- Whether Bernardos filed a complaint with the police or with X.
- When the first AI-made ad using the economist appeared; the earliest dated statement read is the X post of 23 July 2026.
- On which platform the video described in October 2026 circulated.
What the evidence supports
AI involvement: reported. Maldita.es states that it checked all the ad images and that OpenAI's detection tool found the SynthID watermark in each; Maldita.es does not say which tool generated the images; Bernardos says the ads use AI or other tools (Marca) and that the video was created with AI (El Caso). No tool is named and the researcher did not inspect the images or the video. The material depicted Bernardos (depicted_or_impersonated).
6 claims: 6 reported. What the statuses mean
Reported Maldita.es found at least 20 ads that circulated on X on 24 and 25 September 2026 and that appear to involve Bernardos in supposed corruption scenes through AI-generated images; in total they were seen more than 400,000 times.
Causal attribution. Maldita.es reports its own count from browsing X; it calls the figure a minimum.
- maldita.es(opens in new tab) supports · Spanish
'hemos detectado, al menos, 20 anuncios que circularon en X entre el 24 y el 25 de septiembre de 2026 que parecen involucrar al economista español Gonzalo Bernardos en supuestas escenas de corrupción con imágenes generadas con IA'
- maldita.es(opens in new tab) supports · Spanish
'En total, fueron vistos más de 400.000 veces'
Reported According to Maldita.es, one ad image supposedly shows police searching the home of Bernardos, and the ad link could lead to a page imitating El Mundo that says Bernardos was arrested for promoting an investment platform.
Causal attribution. Maldita.es describes the ads and landing pages it inspected; it says the image is not real and that there is no trace of the search having happened.
- maldita.es(opens in new tab) supports · Spanish
'una imagen que supuestamente muestra a la Policía Nacional registrando la casa de Gonzalo Bernardos'
- maldita.es(opens in new tab) supports · Spanish
'puede llevarte a una web que suplanta a El Mundo donde se asegura que Bernardos ha sido “detenido por promocionar una plataforma de inversión”'
Reported Maldita.es says it checked that all the ad images were generated with AI and that OpenAI's tool for identifying whether content was generated with one of its tools detected the SynthID digital watermark in all of them.
Causal attribution. The check is the outlet's own; the images were not inspected by the researcher.
- maldita.es(opens in new tab) supports · Spanish
'hemos comprobado que todas las imágenes de los anuncios han sido generadas con inteligencia artificial'
- maldita.es(opens in new tab) supports · Spanish
'La herramienta de OpenAI para identificar si un contenido ha sido generado con alguna de sus herramientas ha detectado en todas la marca de agua digital de SynthID'
Reported The pages promoted Norden Core, which Maldita.es says is not authorised by the CNMV to provide investment services.
Causal attribution. Maldita.es statement about the regulator's register; the register was not consulted.
- maldita.es(opens in new tab) supports · Spanish
'la supuesta plataforma Norden Core no está autorizada por la Comisión Nacional del Mercado de Valores (CNMV) para prestar servicios de inversión'
Reported Marca reported that on the night of 23 July 2026 Bernardos wrote on X that individuals were using the economist's image to advertise investments and financial products, that all were fake ads using AI or other tools, and asked people to ignore them.
Causal attribution. Bernardos's own statement as quoted by Marca.
- marca.com(opens in new tab) supports · Spanish
'a través de un mensaje en X, publicado en la noche del jueves 23 de julio'
- marca.com(opens in new tab) supports · Spanish
'Desgraciadamente, hay diversos individuos que están utilizando mi imagen para publicitar inversiones, productos financieros o otros temas. Todos son anuncios fake, utilizando la IA u otras herramientas. Por favor, no hacer caso de ellos.'
Reported El Caso reported on 4 October 2026 that Bernardos said on X that many people were asking about a video in which the economist's image announces a supposed financial investment, that the video was created with AI and that Bernardos had never made such promotions.
Causal attribution. Bernardos's own statement as paraphrased by El Caso; the video was not inspected.
- elcaso.elnacional.cat(opens in new tab) supports · Spanish
'son muchas las personas que hablan con él preguntando por un vídeo donde se ve su imagen anunciando una supuesta inversión financiera'
- elcaso.elnacional.cat(opens in new tab) supports · Spanish
'Bernardos ha explicado que el vídeo se ha creado con inteligencia artificial y que él en ningún momento ha hecho promociones de estas características'
Sources
3 sources inspected, from 2 underlying accounts. Sources that repeat one account do not corroborate each other.
- Maldita.es, 1 October 2026: Gonzalo Bernardos con un bolso lleno de billetes: la campaña de anuncios de Twitter (X)(opens in new tab)
s1 · maldita.es · News report · Spanish · Inspected · 1 October 2026 · Primary
- Marca, 24 July 2026: La denuncia Gonzalo Bernardos sobre el uso de su imagen para estafas en internet(opens in new tab)
s2 · marca.com · News report · Spanish · Inspected · Shares an underlying account with another listed source
- El Caso, 4 October 2026: Gonzalo Bernardos, la última víctima de la suplantación de identidad para estafar personas(opens in new tab)
s3 · elcaso.elnacional.cat · News report · Spanish · Inspected · Shares an underlying account with another listed source
How the sources were read, and where the events happened
Spanish-language body read in the original on 2026-10-05 by the research agent (an AI). Passages are quoted verbatim in Spanish. English assertions are the agent's own translation and no human translator reviewed them. Maldita.es investigation, 1 October 2026; its own collection and analysis of the ads. Applies to s1.
Spanish-language body read in the original on 2026-10-05 by the research agent (an AI). Passages are quoted verbatim in Spanish. English assertions are the agent's own translation and no human translator reviewed them. Marca, 24 July 2026; relays the X post of 23 July 2026. Applies to s2.
Spanish-language body read in the original on 2026-10-05 by the research agent (an AI). Passages are quoted verbatim in Spanish. English assertions are the agent's own translation and no human translator reviewed them. El Caso, 4 October 2026; paraphrases a later X post, which was not read directly. Applies to s3.
Event countries: Unknown. Affected-person countries: Unknown. Court countries: Unknown.
The ads ran on X in Spanish and imitated a Spanish newspaper, and Maldita.es locates the advertising accounts in the United Kingdom, Germany, Brazil, Lithuania and Turkey, but no body states where the ads were shown or where Bernardos lives. Maldita.es calls Bernardos a Spanish economist, which is nationality alone. No country is inferred.
Reviewed for publication 2026-10-05: Published under the 2026-09-15 charter as a core case: AI-made images and video depicted a named public figure in fraudulent investment ads and invented arrest scenes, and the person has objected publicly twice. Two reporting chains: an independent Maldita.es analysis of the ads, and the person's own X posts relayed by Marca and El Caso. Each claim stays reported because each rests on one chain.
People described
Gonzalo Bernardos, economist, university professor and television commentator (public figure speaking publicly about the misuse of the economist's own image)
People reported harmed in this case
At least 1 person
0 AI participants · 1 other person harmed
One person, Gonzalo Bernardos, whose image the ads used (Maldita.es; Marca; El Caso). A television presenter shown in one image is not counted because no source carries an account from the presenter. People who saw the ads are not counted because no source reports anyone defrauded. Documented minimum 1.
Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.
Cite this case
Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.
APA
NOPE. (2026). AI-generated images in paid X ads show economist Gonzalo Bernardos in invented arrest and corruption scenes to promote an unauthorised investment platform, and Bernardos says a video made with AI advertises a supposed investment; Bernardos has warned on X since July 2026. AI incidents. https://nope.net/incidents/2026-economist-gonzalo-bernardos-ai-images-video-fake-investment-ads-invented-arrest
BibTeX
@misc{2026_economist_gonzalo_bernardos_ai_images_video_fake_investment_ads_invented_arrest,
title = {AI-generated images in paid X ads show economist Gonzalo Bernardos in invented arrest and corruption scenes to promote an unauthorised investment platform, and Bernardos says a video made with AI advertises a supposed investment; Bernardos has warned on X since July 2026},
author = {NOPE},
year = {2026},
howpublished = {AI incidents},
url = {https://nope.net/incidents/2026-economist-gonzalo-bernardos-ai-images-video-fake-investment-ads-invented-arrest}
} Related cases
Bengaluru: actor Rukmini Vasanth says AI-generated photos passed off as hers are fake and files a complaint after AI-morphed photos and videos of her went viral in May 2026; police arrest three people, reported on 19 and 20 June 2026
According to ETV Bharat Kannada (20 June 2026) and Oneindia Kannada (19 June 2026), Bengaluru police arrested three people accused of using AI technology to morph photographs of the actor Rukmini Vasanth and of circulating AI-made photos and videos depicting her on social media. ETV Bharat reports that the material went viral on social media platforms and in some media in May, that it harmed her dignity and privacy and caused her mental harassment, and that she filed a complaint. In a statement on her social media accounts, quoted by ETV Bharat, she said the AI-generated photos circulating as hers were completely fake and that making and spreading manipulated content was a serious violation of privacy. Police seized three mobile phones and the accused were remanded to judicial custody. Oneindia reports that her face had been obscenely altered with AI technology in a video released on the internet. The reports do not name the tool used.
London: the broadcaster Narinder Kaur says that since sexually explicit images made from her Instagram photographs with a nudification app appeared on X in November 2025 she has been tagged in at least 20 more deepfaked pictures and videos, and that she fell into a depression; no perpetrator has been identified (Daily Mail, 30 September 2026)
Narinder Kaur, a 54-year-old broadcaster and regular TV commentator, told the Daily Mail (30 September 2026) that one morning in November 2025 she opened X and found sexually explicit images of herself, made by taking photographs she had shared on Instagram and removing her clothing with a so-called nudification app. Since then she has been tagged in at least 20 more deepfaked pictures and videos, some linking her to Jeffrey Epstein. She says she fell into a depression in the early weeks and her GP suggested antidepressants, that commenters treated the images as real, and that she fears her teenage children will see them. She has reported each image and had numerous police visits, but no perpetrator has been identified because the accounts are anonymous and deleted when reported. She has joined the Stop AI Abuse campaign.
Jakarta: AI-manipulated TikTok videos showed toll-road businessman Jusuf Hamka and his daughter in detainee uniforms with bribery and corruption narratives that Hamka denies; a TikTok account holder was arrested in November 2025 and is on trial at the Central Jakarta District Court, where Hamka's side refused restorative justice on 29 September 2026
Indonesian businessman Mohamad Jusuf Hamka reported to Jakarta police on 18 October 2025 that videos circulating on TikTok showed him and his daughter as if wearing prosecutors' detainee uniforms, tied to narratives of corruption, bribery and gratification over the Cawang-Pluit toll concession, which he denies and his lawyers call false; his lawyers say the faces were composited into staged scenes with AI, and that other content called him a 'toll mafia' and posted a fabricated poster of him running for president in 2029. Polda Metro Jaya's cyber directorate arrested the holder of the TikTok accounts on 27 November 2025; Hamka and his lawyers say the defendant admitted to police that he made the content on others' instructions for Rp500,000 per item; tvOneNews and TVRI note this is Hamka's side's claim within the contested case, and Poskota reports it is being examined at trial. The case file was reported complete (P21) on 12 August 2026 and the trial opened at the Central Jakarta District Court on 15 September 2026 under the Electronic Information and Transactions Law (Article 35 jo 51(1), maximum 12 years) and alternative counts. Hamka told reporters the content disturbed him and could damage his reputation and that the campaign was systematic. At the 29 September 2026 hearing the judges asked about restorative justice; Hamka's lawyer said there was no room for it unless the defendant named who ordered and paid for the content. The defendant's lawyer said the content related to his work as secretary-general of an anti-corruption committee and that the indictment was unclear.
Maine school district reports teacher targeted by fabricated AI videos and photos
A Maine school superintendent told families that a teacher had been targeted by a malicious digital attack using AI-created videos and photos. The letter, reproduced by WABI, said the content falsely portrayed the teacher and had circulated online, with a small number of students possibly accessing or sharing it. The district said the hosting account had been reported and the video taken down, and offered support to the teacher. The first creation date and specific content are unknown.
If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.