AI is undermining democracy, Oireachtas committee to hear
Eileen Culloty, an associate professor of communications at DCU and head of the Irish hub of the European Digital Media Observatory, will warn the Oireachtas AI committee that we may be repeating errors of the 1990s regarding tech deregulation. Picture: DCU
Information derived via artificial intelligence (AI) is “fundamentally unreliable”, while the prevalence of such data is ruining people’s ability to reason effectively, an Oireachtas committee will be told tomorrow, Tuesday.
TDs and senators on the Oireachtas AI committee are set to hear that society is making “the same mistake again” in terms of a similar situation in the 1990s, when tech deregulation became the norm in order to support innovation.
Eileen Culloty, an associate professor of communications at DCU and head of the Irish hub of the European Digital Media Observatory, will tell the committee that, while the world is “still dealing with the consequences” of such lax regulation at the dawn of the internet age, the rush by governments to hasten developments in AI is likely to cause similar issues for generations to come.
“Faced with regulatory choices, it is striking that we appear to be making the same mistake again,” she will tell the committee.
Representatives from media group Newsbrands Ireland and the National Union of Journalists will be before the committee to discuss truth and democracy amid the advent of AI.
Ms Culloty will tell committee members that AI, though having many legitimate potential uses, is not designed to tell the truth when questioned, but rather to return an answer which may be correct.
“Regarding truth, generative AI is fundamentally unreliable. It does not aim at factual accuracy,” she will say.
“It offers outputs that are statistically plausible based on its training data.
“Tech companies overhype AI systems and embed them into everyday life.
“We are offered convenient, but unreliable, tools to generate, summarise, and evaluate content,” she will say, adding that such a glut of offerings “undermines our capacity to reason”.
That fact, Ms Culloty will argue, represents a “deeper issue” than that of the world being flooded with AI slop and scams:
Separately, Newsbrands Ireland chair Sammi Bourke will tell the committee that while news publishers are not anti-AI, the use by artificial intelligence portals of online journalism without payment has left traditional quality journalism “under existential threat”.
She will say that news publishers worldwide “are sounding the alarm about the unchecked use of copyrighted material in AI training”.
Typically AI models are ‘trained’ on data freely available on the internet, including news reports.
However, relying on such material means that AI prompts will suffer from the same inaccuracies as any included in such journalism.
Ms Bourke will argue that should journalism become unsustainable then “the consequences will be weaker scrutiny of power, greater exposure to disinformation and misinformation, and declining public trust in democratic institutions”.
“The actions and decisions taken now will determine whether AI strengthens our democratic institutions or quietly destroys them” she will say.
The fact that neither the media nor creative sectors were consulted by the National AI Advisory Council in compiling its advice paper on the topic is “remarkable”, Ms Bourke will also say.
“Maybe this is not surprising when you consider that council membership, which includes representatives from AI companies, has no representation from the media or the arts,” she will say, while noting that the council did nevertheless acknowledge that “our copyright framework is under strain from rapid AI development and that gaps and ambiguities”.
CONNECT WITH US TODAY
Be the first to know the latest news and updates
