At an Oireachtas hearing on AI today, Electoral Commission CEO Art O’Leary warned about AI deepfakes, while other experts said it was facilitating the rise of the “far right”
Speaking to the Joint Oireachtas Committee on Artificial Intelligence, An Coimisiún Toghcháin chief executive, Art O’Leary, described the AI deepfake “challenge” as “vivid” and said that the recent Catherine Connolly deepfake clip, first reported by the Irish Independent, “showed how easily such content can be produced and how often we may see similar attempts”.
The Oireachtas Committee was also told by UK expert Dr Dan McQuillan that AI is “toxic to democracy” and is being used to strengthen far right political “reactionaries”, as well as seducing students into replacing “critical thinking” with “slop”.
“What do we talk about when we talk about Al? We’re talking about chatbots and deepfakes, synthetic text, voices and images, about extraordinary efficiencies and equally extraordinary vulnerabilities,” said O’Leary.
“We’re talking about how political actors might use Al and how malign actors might weaponise it to distort debate or undermine confidence.”
Asked whether he had used AI to help prepare his speech to the Oireachtas Committee, O’Leary said he had not.
“Al undermines the ideals of truth and democracy, said Dr McQuillan, a lecturer in computing at the University of London.
“Al has an adversarial relation with the truth. The core of its calculations are correlations, not causal relations, so its outputs are plausible rather than factual. Al’s pattern recognition is a form of computational conspiracy theory and its outputs are disinformation even when they appear to be accurate.
“Large language models are sold as learning accelerators, but actually substitute slop for critical thinking,” he said.
“Unfortunately, they are becoming the first port of call for young people from everything from essays to relationship advice. In this context, Al undermines the replenishment of a citizenry with the capacity for independent thought. In its systemic effects, Al will fail to solve problems, cause collateral damage and benefit reactionary politics.
“I suggest the Committee be clear with itself what it’s endorsing when it endorses Al. At best, the alleged benefits to health care or education really amount to algorithmic Thatcherism. A more likely outcome is that widespread AI adoption will strengthen the far right.”
