Incidents of scams, fraud, and harassment are on the rise as the use of “deepfakes” becomes more widespread and advanced.
Deepfakes involve modifying existing images or videos using artificial intelligence (AI) technology, reports RTE.
A notable example includes the image of Pope Francis wearing a white puffer jacket — a convincing deepfake. The technology is rapidly improving, especially in creating realistic videos and audio.
While some deepfakes are created for harmless or creative reasons, others have more dangerous implications, reports RTE.
“People might use it for entertainment, for creativity, or just to create low cost material,” explained Dr Eileen Culloty from the DCU School of Communications.
“But the flip side is manipulation, scams and fraud but also bullying and harassment, and in particular sexual harassment using deepfakes,” she added, reports RTE.
Deepfakes have also been linked to attempts at political disruption. In 2022, a fake video showing Ukrainian President Volodymyr Zelensky falsely surrendering was circulated online.
Fake videos have also misused real footage from RTÉ News. These altered clips are being posted on social media and gaming platforms, reports RTE.
Presenters’ voices are being manipulated and fake interviews inserted into news segments to give the impression that the news outlet is endorsing what is, in reality, a scam.
As these fakes become more realistic, they are harder to detect. Currently, there is limited legal protection in this area, though legislation is in development, reports RTE.
Barry Scannell, an expert in AI and tech law, said victims of manipulated audio or visuals face significant legal obstacles.
“Without any laws on deepfakes, the victims of deepfake technology are having to be as creative, if not more creative, than the perpetrators using this technology,” he said, reports RTE.
There are some legal safeguards in place for brand misuse, under the concept of ‘passing off’, but this doesn’t extend to individuals’ images in the same way.
“Ireland doesn’t have what’s called a ‘personality right’, like they have in the United States. Where you actually have a right to your image, we don’t have that,” he added, reports RTE.
Mr Scannell, a member of the AI Advisory Council, noted that the group has advised the Government to introduce such a legal right, reports RTE.
The forthcoming AI Act contains regulations on deepfakes, but it will not be implemented until August next year.
“The AI Act only says that you need to mark it as a deepfake, provided that it isn’t obvious,” Mr Scannell said, reports RTE.
“The problem is that deepfakes are now so sophisticated that what’s obviously a deepfake isn’t clear,” he added, reports RTE.
Online platforms offer tools for reporting deepfakes, but often those complaints are dismissed for not violating platform rules.
RTÉ News shared one such deepfake video that had been on Facebook for over a year with Meta.
In response, the social media company stated: “We are constantly developing new ways to make it harder for scammers to deceive others – including using facial recognition technology…we have removed and disabled the video that was shared with us,” reports RTE.
With more people getting their news online, the ability to question content and verify sources is becoming an essential skill for separating fact from fiction.
Mr Scannell said: “There isn’t any silver bullet. One of the things we should very much consider, and that I would strongly favour, would be teaching critical thinking in schools,” reports RTE.
Tell us your thoughts in the Facebook post and share this with your friends.
