Irish Times falls victim to a hoax.
The Irish Times newspaper has apologized for running a hoax article created using artificial intelligence. It said it was a breach of trust and a deliberate deception.
The respected daily ran the fake article on Thursday. It was titled “Irish women’s obsession with fake tan is problematic.” It became the second most-read item on the newspaper’s website.
Bosses removed the article on Friday. They said on Sunday they were “genuinely sorry.”
How the hoax was done
The hoax article was submitted under the fictitious Ecuadorian writer Adriana Acosta-Cortez’s byline. It argued that the use of fake tan by Irish women was an issue of cultural appropriation.
An individual controlling a Twitter account under the same name said they had used the AI language tool GPT-4 to create most of the article. They also used the image generator Dalle-E 2 to create a profile picture for the opinion piece.
The hoaxer said they were an Irish college student who wanted to create debate around identity politics and “give my friends a laugh.”
The Irish Times said the author had “engaged with the relevant editorial desk.” They took suggestions for edits, offered personal anecdotes, and supplied links to research.
The gap in pre-publication procedures
The editor of the Irish Times, Ruadhan Mac Cormaic, wrote a statement on Sunday. He said the publication “was a breach of the trust between the Irish Times and its readers.”
He said the incident had highlighted a gap in their pre-publication procedures.
Underlined one of the challenges raised by generative AI for news organizations. However, he said they would learn and adapt.
He did not reveal how the hoax was discovered.
AI challenges for news organizations
Artificial intelligence is increasingly used to create content for various purposes. Some are benign, such as generating captions for images or summaries for texts. Others are malicious, such as spreading misinformation or propaganda.
News organizations need to be vigilant and verify the sources and authenticity of their content. They also need to be transparent and accountable to their readers.
The Irish Times is not the first news outlet to be duped by AI-generated content. In 2019, The Guardian published an op-ed by an AI system called GPT-3. However, it was clearly labeled as such and done as an experiment.
The Irish Times said it would work to prevent such hoaxes from happening again.