RTÉ presenter Miriam O'Callaghan is weighing a second legal challenge against Meta Platforms Ireland after AI-generated images depicting her with severe facial injuries appeared on social media. The images, which show her bruised and bloodied, have reignited concerns over the use of artificial intelligence to create defamatory content and the legal obligations of platforms to remove it.
O'Callaghan previously settled a case against Facebook in 2022 over fake skincare ads that used her likeness without consent. That settlement included an unreserved apology and a commitment from Meta to create a dedicated scam ad reporting tool for Irish users.
"When you see a picture like that, where you're looking like you're being arrested because you've done something wrong, it makes you think, oh my goodness, how can that be up online?" O'Callaghan said.
Previous legal action against Facebook
In 2022, Miriam O'Callaghan received an unreserved apology from Facebook, now Meta Platforms Ireland, after a five-year battle over misleading advertisements. The fake ads used her image to promote skincare products without her permission, causing distress and reputational damage.
As part of the settlement, Meta agreed to establish an additional scam ad reporting tool that lets Irish users submit more details about misleading adverts to a specialist team for review. Speaking after the settlement, O'Callaghan described it as "a good day" following the protracted dispute.
AI-generated violence raises the stakes
The latest images are more disturbing than the earlier fakes. "In that picture, I looked like I'd been seriously beaten up," O'Callaghan told the Sunday Independent. "You just wonder the mindset of the person who decides this is the correct clickbait, to do that to my face."
Her lawyer, Paul Tweed, previously traced the 2022 scam ads to operators in eastern European countries. O'Callaghan said those behind the new images are "incredibly experienced" and often work for large companies that target well-known individuals. "They take the image of someone very well known in their own country and they mock it up like that. But this is a different level, because of the bruising and the blood."
She has not yet decided whether to launch fresh legal proceedings. "I'm not going to sit idly by and let them continue, because once you allow it they multiply and multiply," she said. "They decide, oh, this person is worth doing this to because we're getting a lot of clicks, and then it spreads like wildfire." O'Callaghan added that she will monitor the situation and contact Facebook and X again.
For legal professionals, cases like O'Callaghan's illustrate the growing need to understand AI-generated content and its implications for defamation and privacy law-a focus area in AI for Legal Professionals Courses.
Why this matters for legal professionals
O'Callaghan's experience highlights how AI tools can generate realistic, harmful imagery that existing legal frameworks struggle to address. The cross-border nature of the operations complicates jurisdiction and enforcement, while platforms' content moderation systems often fail to catch such material before it spreads. Lawyers advising clients on reputation management or platform liability need to grasp both the technical capabilities of AI and the evolving regulatory response. The case also signals potential growth in litigation against social media companies over AI-generated harms, creating demand for legal expertise in this emerging area.
Your membership also unlocks:
