Ireland ought to use its upcoming presidency of the Council of the European Union to ban artificial intelligence (AI) models which generate intimate images and child sex abuse material, an expert body has said.
The State’s Artificial Intelligence Advisory Council, which advises the Government, has warned that “high-velocity, automated abuse” associated with models such as X’s Grok is likely to become “increasingly common”.
It also said many people who may have been generating so-called “nudification” images using Grok, the AI model owned by Elon Musk‘s social media giant X, may not be aware they could be breaking the law.
On Thursday, X said the Grok AI function would no longer allow users to manipulate photos of people to appear in revealing clothing in places where such actions were illegal.
X bowed to pressure following deepening controversy and safety concerns from UK regulators.
Minister of State Niamh Smyth is to meet X executives on Friday and has said she will tell them that what happened “over the last week is illegal and they will face the full rigours of the law here in Ireland in doing so”.
On Thursday she welcomed moves by the social media platform to further restrict the controversial Grok app.
She said Ireland had “very robust laws in place” to deal with Grok and similar applications and said these laws “make it very clear … it is illegal in this country to create these images”.
“It is illegal in this country to disseminate these images. And it is illegal in this country to share these images,” she said.
She said she had urged “swift, co-ordinated action with the EU Commission, including a formal investigation and interim measures to protect citizens across all 27 member states”.
The Commission, which is the EU’s executive arm, has taken investigative steps in relation to X and its obligations under the Digital Services Act, Ms Smyth said.
It will now assess the changes to Grok announced by X to ensure these “effectively protect citizens of the EU”, she said.
The AI Advisory Council published a policy paper on Thursday in response to allegations that non-consensual intimate images had been generated and shared by Grok.
The organisation said the Government should “use the Irish EU presidency to work with the other EU member states” to change EU-wide legislation to “include the prohibition of AI practices which allow users to generate nonconsensual intimate images and child sex abuse material”.
The advisory group also urged the Government to publish guidance explaining how potential victims of image-based sexual abuse can report the potential crime and preserve evidence.
It said there should be a publicly funded advertising campaign “explaining to the public that using AI to create nonconsensual intimate images of people which are automatically shared is a potential criminal offence”.
The council also said Ireland ought to standardise the policies around “AI-enabled harms” to design “effective responses to high-velocity, automated abuse which is likely to become increasingly common”.