Skip to main content
NOPE
Low reported severity Media Coverage

Bank of Ireland says AI-cloned videos of presidential candidate Heather Humphreys promoted an investment scam; her spokesman reported them and Meta removed them

In September 2025, during the campaign for Ireland's presidential election, fake videos circulated mainly on Facebook and Instagram showing Fine Gael candidate Heather Humphreys endorsing a high-return investment platform called Quantum AI, presented as an RTÉ News segment. Bank of Ireland warned consumers on 11 September that the content was entirely fabricated using AI-driven voice and image cloning and that more such adverts were likely, without stating how it reached that view. A spokesman for Humphreys said the videos had been reported to Meta and called for their immediate removal, and Meta said that evening that it had taken them down. No person who lost money to these particular videos is described in the sources.

AI system
Unidentified video tool
Occurred
Sept 2025
Reported
11 September 2025
Event location
Ireland
What the AI did
Depicted or impersonated the person
Reported harm
Reputational Harm
Whose AI use
Someone else’s AI use
Setting
Finance · Public life
Evidence
AI involvement reported · Causal attribution supported · 3 sources
6 claims: 6 reported. 4 open questions
People reported harmed
1 person

AI system as recorded: Unidentified voice and image cloning tool used to make fake investment videos of an Irish presidential candidate, as described in a Bank of Ireland warning

What Happened

The videos. TheJournal.ie reported on 11 September 2025 that fake videos had circulated in the preceding days showing Heather Humphreys, then Fine Gael's candidate for president, endorsing an investment platform called Quantum AI in what was presented as an RTÉ News segment. Bank of Ireland said they circulated primarily on Instagram and Facebook.

The AI attribution. Bank of Ireland's warning, relayed by the Connaught Telegraph, says the video falsely portrays her endorsing a high-return investment scheme, with the content entirely fabricated using AI-driven voice and image cloning. The bank's head of fraud called the videos highly convincing and designed to lure people into fraudulent schemes. The warning does not say how the bank established that the videos were AI-made. In October 2025 Investigate Europe described a Facebook post in which she appears to introduce Quantum AI, stated that she never said this, and described such videos as AI-generated fakes made with voice-cloning software. The Central Bank of Ireland had warned in 2024 that Quantum AI uses AI-generated deepfake videos of high-profile people (TheJournal.ie).

Her response and the takedown. A spokesman for Humphreys said the fake videos had been reported to Meta and called for them to be removed immediately, and encouraged voters to check what they see online. A Meta spokesperson said the videos had been taken down and that such adverts breach its policies.

Limits. The sources do not identify who made the videos or which tool was used, how many people saw them, or anyone who invested after seeing them. The videos themselves were not inspected.

Reported harm

Fake investment adverts that Bank of Ireland says used AI voice and image cloning showed a presidential candidate endorsing a scam platform; her spokesman reported them to Meta and demanded their removal, and Meta took them down (reported).

Outcome

Unknown

Her spokesman reported the videos to Meta and asked for their immediate removal; a Meta spokesperson said on 11 September 2025 that the videos had been taken down (TheJournal.ie). Bank of Ireland warned consumers to expect more such adverts. No police complaint or prosecution is reported.

What remains unknown

  • Who made the videos and which AI tool was used.
  • How Bank of Ireland established that the videos were AI-made.
  • How many people saw the videos and whether anyone invested after seeing them.
  • Whether a complaint was made to the Gardaí.

What the evidence supports

AI involvement: reported. Bank of Ireland's consumer warning, as relayed by the Connaught Telegraph (the AI wording is in the warning's text, not in the head of fraud's quoted remarks), says the video showing the candidate endorsing a high-return investment scheme was entirely fabricated using AI-driven voice and image cloning; Investigate Europe also describes such videos as AI-generated fakes made with voice-cloning software. The cloned likeness is what the adverts presented to viewers and what her spokesman reported to Meta. The bank does not state its technical basis, and the tool and maker are not identified.

6 claims: 6 reported. What the statuses mean

Reported Fake videos circulating mainly on Instagram and Facebook in September 2025 showed presidential candidate Heather Humphreys endorsing an investment platform, Quantum AI, in a purported RTÉ News segment.

Causal attribution. Description of the adverts by Bank of Ireland and TheJournal.ie; the adverts were not inspected. Both sources rest on the bank's warning for the circulation details.

  • con-telegraph.ie(opens in new tab) supports · English
    'Fraudsters are already using deepfake technology to clone the image and voice of presidential election candidate Heather Humphreys in fake investment videos circulating online.'; 'Bank of Ireland is warning consumers to watch out for more of these adverts, mainly originating via Meta platforms, over the coming weeks.'
  • thejournal.ie(opens in new tab) supports · English
    'The videos, which have circulated in recent days, cloned the Fine Gael politician’s image and voice using deepfake technology.'; 'The footage falsely portrayed an RTÉ News segment in which ‘Humphreys’ endorsed an investment platform, Quantum AI.'; 'The fake videos of Humphreys have primarily circulated Meta social media platforms Instagram and Facebook, Bank of Ireland said.'
Reported Bank of Ireland said the video was entirely fabricated using AI-driven voice and image cloning and that the videos were highly convincing and designed to lure people into fraudulent schemes.

Causal attribution. The bank's own attribution, relayed by the Connaught Telegraph; the AI sentence is unquoted text of the relayed warning, the head of fraud's quoted remarks do not mention AI, and the technical basis is not stated.

  • con-telegraph.ie(opens in new tab) supports · English
    'The video falsely portrays Heather Humphreys endorsing a high-return investment scheme, with the content entirely fabricated using AI-driven voice and image cloning.'; '"These videos are highly convincing, designed to lure unsuspecting individuals into fraudulent schemes.'
Reported A spokesman for Humphreys said the fake videos had been reported to Meta and called for their immediate removal.

Causal attribution. The candidate's spokesman's statement as reported by TheJournal.ie.

  • thejournal.ie(opens in new tab) supports · English
    'A spokesman for Humphreys said today that the fake videos had been reported to Meta, and called for them to be removed from the social media platforms “immediately”.'
Reported A Meta spokesperson said the videos had been taken down.

Causal attribution. Meta's statement as reported by TheJournal.ie; Meta did not describe how the videos were made.

Reported Investigate Europe stated that Humphreys never said what a Facebook video showed her saying about Quantum AI and described such videos as AI-generated fakes made with voice-cloning software.

Causal attribution. Investigate Europe's own reporting in October 2025; it does not say which post it examined or how it established the AI generation.

  • investigate-europe.eu(opens in new tab) supports · English
    'Except Humphreys, and Pistorious, never said any of this. The videos are a part of a growing wave of elaborate AI-generated fakes created with advancing voice-cloning software'
Reported The Central Bank of Ireland had warned the previous year that Quantum AI uses AI-generated deepfake videos and photos of high-profile people on social media.

Causal attribution. Context on the scheme from TheJournal.ie's summary of a regulator's earlier warning, not about these particular videos.

  • thejournal.ie(opens in new tab) supports · English
    'The Central Bank of Ireland warned last year that Quantum AI claims to provide cryptocurrency trading services using AI technology, and uses AI-generated deepfake videos, fake newspaper articles and photos of high-profile people through social media'

Sources

3 sources inspected. Sources that repeat one account do not corroborate each other.

How the sources were read, and where the events happened

English original read in full; no translation. Applies to s1, s2, s3.

Event countries: Ireland. Affected-person countries: Ireland. Court countries: Unknown.

Bank of Ireland warned consumers about adverts using an Irish presidential candidate (Connaught Telegraph), the videos presented a purported RTÉ News segment (TheJournal.ie), and the Facebook post described by Investigate Europe pitched the platform to Irish families. She is a candidate for the Irish presidency. No court is involved.

Reviewed for publication 2026-10-29: Published because a bank's fraud function attributed the fake investment videos to AI voice and image cloning and the depicted candidate's spokesman reported them to Meta and demanded their removal. The AI attribution is the bank's, with no technical basis stated.

People described

A presidential candidate and former government minister in Ireland whose image and voice were used in fake investment adverts

People reported harmed in this case

1 person

0 AI participants · 1 other person harmed

One person: the presidential candidate whose image and voice the videos used, whose spokesman reported them to Meta and called for their removal (TheJournal.ie). No deceived investor is described in the sources; viewers and the consumers warned by the bank are not counted.

Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.

Tags

#deepfake#investment-scam#voice-cloning#election#meta

Cite this case

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Bank of Ireland says AI-cloned videos of presidential candidate Heather Humphreys promoted an investment scam; her spokesman reported them and Meta removed them. AI incidents. https://nope.net/incidents/2025-ireland-ai-cloned-investment-videos-of-presidential-candidate-removed-by-meta

BibTeX

@misc{2025_ireland_ai_cloned_investment_videos_of_presidential_candidate_removed_by_meta,
  title = {Bank of Ireland says AI-cloned videos of presidential candidate Heather Humphreys promoted an investment scam; her spokesman reported them and Meta removed them},
  author = {NOPE},
  year = {2026},
  howpublished = {AI incidents},
  url = {https://nope.net/incidents/2025-ireland-ai-cloned-investment-videos-of-presidential-candidate-removed-by-meta}
}

Related cases

Low Unidentified video tool

Fact-checkers say videos impersonating an Albanian cardiologist and a journalist to sell hypertension pills were likely AI-made; both men disowned them

A Facebook page in the name of the Albanian cardiologist Dr. Spiro Qirko published a video claiming he had discovered a cure for hypertension, and a second video presented as an interview of him by the journalist Ilir Topi, promoting a product called Hyper Caps. Kallxo's fact-checking unit Krypometer, in an article of 30 March 2026, reports that it ran the video through the AI-content detector Hive Moderation, which gave a 96.4% probability that the video and the speech were AI-generated, and that Hyper Caps is not on the Kosovo medicines agency's list of products authorised for marketing. Both men told Krypometer they had nothing to do with the page or the video. The journalist said it was AI-generated and fake, and the cardiologist said he felt deeply offended by the fraud and had never recommended herbal remedies for hypertension. Who made the videos, how widely they circulated and whether anyone bought the product are not reported.

Medium Unidentified video tool

Spain: Partido Popular AI video showed PSOE figures and the prime minister's wife confessing alleged crimes; Leire Díez and Begoña Gómez file complaints

During the campaign for Spain's 29 November general election, the Partido Popular (PP) posted on its official social media accounts on 21 October 2026 a video made with artificial intelligence, labelled as AI content, that parodies a PSOE campaign video. It shows AI-generated likenesses of the prime minister's wife Begoña Gómez, former prime minister José Luis Rodríguez Zapatero, former minister José Luis Ábalos, former PSOE officials Santos Cerdán and Montse Mínguez and former party member Leire Díez, with imitated voices, each appearing to confess alleged crimes or irregular conduct. PSOE sources told Europa Press the party was studying legal action over the video, which elDiario.es and Cadena SER describe as impersonating these people, and PSOE's organisation secretary said the PP had nothing to offer but insults. PP spokespeople defended the video, saying it was clearly recognisable as AI. Europa Press reported more than a million views on X. By 23 October 2026 Leire Díez, one of the people depicted, had filed a criminal complaint (querella) against the PP for alleged defamation with publicity, offences against moral integrity and digital identity impersonation, asking for the video's provisional removal; Europa Press reports it was sent to the Madrid courts at Plaza de Castilla and El País that it was filed at the Audiencia Nacional. The complaint says the video uses artificial intelligence and her image and voice to accuse her of a serious crime, that its statements are false, that it had more than 21 million views, and that she and her daughters have had to isolate at home and she suffered an anxiety attack. Jurists consulted by El País considered a criminal case unlikely to succeed given the latitude for political satire and the video's AI label. On 28 October 2026 the prime minister said in a podcast interview that his wife, Begoña Gómez, had filed a complaint over the video for impersonation of her image; RTVE and El Periódico, which say they had access to the filing, report that it is directed at whoever runs the PP's official social-media profiles and anyone in the party who instructed them, asks that PP secretary general Miguel Tellado testify, and says she never said those words, took no part in the video and did not consent to the use of her image or voice. The PP answered on social media with irony.

Low Unidentified video tool

Mississippi Senate candidate posted undisclosed AI video of Sen. Hyde-Smith fleeing reporters; campaign deleted it after AG said it appeared to break state law

In early October 2026 the campaign of Scott Colom, the Democratic candidate for the U.S. Senate in Mississippi, posted on social media a 19-second AI-generated video showing a woman made to look like his opponent, Republican Sen. Cindy Hyde-Smith, running from a crowd of reporters at the U.S. Capitol and getting into a car. Local outlets report that the post called it "based on a true story" and did not disclose that it was made with AI. Hyde-Smith's campaign manager called it an AI video and part of a campaign of "lies and deception". On 22 October the Mississippi Republican Party asked the state Attorney General to investigate whether the video broke a state law against knowingly spreading a deceptive AI depiction of a candidate within 90 days of an election, and to seek an injunction. Colom's campaign manager defended the campaign's use of AI and called the complaint desperate; the law exempts satire and parody. On Friday 23 October Attorney General Lynn Fitch said she had notified the campaign that its AI-generated ads appeared to violate the state law against political deepfakes and asked that they be removed that afternoon, or her office might seek an injunction. Colom called the video parody but said the campaign would delete the post rather than spend donors' money on litigation, and by that afternoon two Mississippi stations reported that it no longer appeared on the campaign's Facebook page. No court has ruled on whether the video broke the law.

Low Unidentified video tool

North Carolina attack ad used AI-generated video of congressional candidate Jamie Ager dancing with drag performers, fact check says; performers object

In October 2026 a 30-second television attack ad paid for by the National Republican Congressional Committee (NRCC) and approved by Republican state Rep. Jennifer Balkcom, the candidate in North Carolina's 11th congressional district, showed her Democratic opponent Jamie Ager at a barn dance with drag performers. A News & Observer fact check says the ad features videos generated by artificial intelligence and appears to manipulate real footage from one of Ager's own campaign ads, adding a pride flag to the barn wall and filling the barn with drag queens; the ad carries no AI disclaimer, and North Carolina has no AI-disclosure law for political ads. An Asheville drag performer shown in the ad told WLOS that the scene never happened, that she has never met Ager, that her likeness was used without permission and that she sent a cease-and-desist letter; Ager called the ad 100% fake AI-generated lies. Other performers named in the ad also objected. The NRCC said it stands by the ad.

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.

Back to all incidents