Artificial intelligence (AI) apps that “nudify” images of people without their consent are not currently banned under European Union regulations.
The European Commission confirmed in response to a question from a number of MEPs, including Fianna Fáil’s Barry Andrews, that even AI apps that generate child sex abuse imagery are not banned under the existing AI law.
Earlier this year, the commission began an investigation into social network X under the Digital Services Act after the company’s AI chatbot Grok was used to generate non-consensual, intimate images – including some alleged to be child sex abuse.
If the investigation finds X breached existing rules on illegal content, it could face a fine of up to 6 per cent of its total worldwide annual turnover.
READ MORE
The controversy has, however, prompted calls for stricter laws that explicitly ban AI tools used to generate so-called deep fakes – artificially created but convincing digital content, including images and videos.
A group of more than 50 MEPs have urged the European Commission to explicitly ban “AI-powered nudity and undressing applications from the EU market”.
They had asked the commission to confirm if such apps were “clearly banned” from the EU market under the AI Act or any other EU legislation.
In a response this week, the commission said it had examined whether the current AI Act bans systems “that generate child sexual abuse material or sexually explicit deepfake nudes, and concluded that they would not”.
“The commission is aware that member states and members of the European Parliament are proposing to strengthen the current protection and include such AI systems in the context of the AI Omnibus negotiations.”
[ How Ireland’s regulators are taking action against Big TechOpens in new window ]
The 2024 AI Act marked the first time the commission tried to regulate AI. The EU is trying to streamline the same law with an upcoming AI Act Omnibus.
The commission has been accused of watering down some of its AI laws, following a push back from big tech companies and the US government.
“That such nudify apps like Grok are not banned in the EU is shocking but shows how difficult it is to regulate such ever-changing AI technology,” Andrews said.
“There is one opportunity, the ongoing talks between the commission, [European] Parliament and national governments right now on simplifying the AI Act. We must all come together urgently, set aside politics, and agree that these horrible apps are banned from Europe as soon as possible.
“Tech companies who allow such apps online in future must see their services rapidly suspended.”
[ Number of Garda investigations into Grok AI images rises to 244Opens in new window ]
Meanwhile, Jackie Fox, the mother of Nicole Fox Fenlon, has called for the law passed in her daughter’s memory to include a new ban on AI-generated intimate images.
Fox, who was addressing the European Parliament in a speech to mark International Women’s Day, successfully campaigned for Coco’s Law after her daughter Nicole, known by the nickname Coco, died by suicide in 2018 aged 21 after relentless online abuse.
The resulting 2020 legislation banned the sharing of intimate images without consent, as well as threatening and harmful communications.
Fox said the law needed to be amended to include AI-generated intimate images.
“There needs to be a law to prosecute people [who think] it’s okay to share an image of someone, or to change or to put someone’s face in that image, and remove their clothes, or whatever they may do,” she said.
“It’s absolutely soul-destroying for the people that they do this to. And if there’s not anything in place, it’ll just get bigger and bigger and worse and worse.”
