{"id":489792,"date":"2026-06-09T00:20:16","date_gmt":"2026-06-09T00:20:16","guid":{"rendered":"https:\/\/www.newsbeep.com\/ie\/489792\/"},"modified":"2026-06-09T00:20:16","modified_gmt":"2026-06-09T00:20:16","slug":"research-raises-concerns-about-impact-on-justice-systems-of-using-ai-the-irish-times","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/ie\/489792\/","title":{"rendered":"Research raises concerns about impact on justice systems of using AI \u2013 The Irish Times"},"content":{"rendered":"

New Irish research has raised concerns about the impact of the use of Artificial Intelligence<\/a> on access to justice and on fairness and reliability in justice systems worldwide.<\/p>\n

Having examined whether an AI judge can \u201cbe talked into a verdict\u201d, the study, published on Monday, found the answer to that question is \u201cyes\u201d, and \u201csometimes to a striking degree\u201d.<\/p>\n

Persuadability and LLMs as Legal Decision Tools, by Oisin Suttle of Maynooth University\u2019s School of Law and Criminology and David Lillis of UCD\u2019s School of Computer Science, is described as the \u201cfirst systematic study\u201d of how persuadable AI models are when confronted with hard legal questions from real cases.<\/p>\n

It is published as the Irish judiciary is preparing new guidance for practitioners and parties on AI use. <\/p>\n

\u201cIt is important that a judge should be open to persuasion. Otherwise, there would be no point in letting the parties argue their case,\u201d said Suttle. <\/p>\n

\u201cHowever, we don\u2019t want a judge to be too persuadable. They should decide on the law and the facts of the case, rather than being swayed by an especially skilled advocate.\u201d<\/p>\n

The study results show that AI models in legal settings can be \u201cvery persuadable\u201d and that raises worries about both reliability and fairness, he said.<\/p>\n

It is \u201cessential that society and decision-makers don\u2019t simply follow hype when it comes to the capabilities of modern AI systems\u201d, said Lillis, stressing the importance of scientific investigation of their capabilities and characteristics \u201cparticularly in sensitive areas such as legal decision-making\u201d. <\/p>\n

The research examined the extent to which LLMs (Large Language Models) in decision support roles are subject to persuasion in legal settings. <\/p>\n

[\u00a0Irish teenagers as young as 13 worried about career impact of artificial intelligenceOpens in new window<\/a>\u00a0]<\/p>\n

\"AIAI in courts could deepen inequality in access to justice, study warns. Photograph: iStock <\/p>\n

Frontier AI models, it found, are \u201csignificantly influenced\u201d by the quality of legal arguments presented to them. <\/p>\n

The researchers designed an experiment using real appellate court cases drawn from Ireland, England and Wales, and the United States. They pitted AI models against each other in a simulated legal contest with one AI acting as an \u201cAdvocate\u201d arguing one side of the case, another arguing the other side, and a third playing the role of \u201cJudge.\u201d<\/p>\n

Having varied which AI model argued each side and measuring how often each \u201cwon,\u201d the team quantified how much argument quality shifted the outcome. <\/p>\n

Every AI model tested was \u201cmeasurably persuadable\u201d, the study found. <\/p>\n

Across 20 different model configurations, including leading systems from Anthropic, Google, OpenAI, Mistral and DeepSeek, stronger advocate models won between 58 per cent and 71 per cent of cases on average. In the \u201cmost extreme\u201d matchups, the stronger advocate won more than 90 per cent of the time.<\/p>\n

Noting that administrative agencies, courts, and tribunals in multiple jurisdictions are piloting AI systems to handle everything from research to case triage and judgment drafting, the researchers argue that persuadability is a dimension of these systems that has received almost no attention until now.<\/p>\n

The implications are \u201cparticularly acute for access to justice\u201d, they say. <\/p>\n

If an AI decision-maker is heavily influenced by the quality of legal argumentation, parties with access to sophisticated legal representation or more powerful AI advocates \u201cwill have a systematic advantage over those without\u201d.<\/p>\n

The study found larger, more capable AI models tended to be \u201csomewhat less persuadable\u201d than smaller models but cautioned against interpreting that as reassurance. <\/p>\n

Even the least persuadable models showed \u201csignificant sensitivity\u201d to argument quality, the researchers said. <\/p>\n

The reasons for lower persuadability in smaller models also appeared, \u201ccrucially\u201d, to be different. Rather than exercising independent judgment, \u201cthese models may simply struggle to evaluate competing arguments at all\u201d.<\/p>\n

While not saying AI should not be used in legal settings, the study argues persuadability must be measured and disclosed as a standard part of evaluating any AI legal decision tool.<\/p>\n

Persuadability and LLMs as Legal Decision Tools, available here<\/a>, was presented on Monday at the 21st International Conference on Artificial Intelligence and Law in Singapore.<\/p>\n

[\u00a0\u2018More harmful than helpful\u2019: Young people sour on AIOpens in new window<\/a>\u00a0]<\/p>\n","protected":false},"excerpt":{"rendered":"New Irish research has raised concerns about the impact of the use of Artificial Intelligence on access to…\n","protected":false},"author":2,"featured_media":489793,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[220,218,219,61,60,80],"class_list":["post-489792","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-ie","tag-ireland","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/489792","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/comments?post=489792"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/489792\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media\/489793"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media?parent=489792"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/categories?post=489792"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/tags?post=489792"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}