Four out of five Irish companies say responsible AI is not a priority, while an overwhelming majority are unprepared for the next EU AI law, new research published on Tuesday reveals.
Responsible AI is defined as ensuring that AI is useful and innovative, as well as safe, secure, fair, transparent, and accountable. According to the PwC Ireland Responsible AI Survey, the majority of Irish respondents (77%) say they have started working towards responsible AI and AI governance practices, but less than a fifth (19%) say it is a business priority across their organization, and almost two-thirds (65%) of Irish respondents say responsible AI and AI governance practices are a priority. respondents say insufficient resources are allocated to governance.
PwC Ireland Chief Operating Officer David Lee said: “This research highlights that the responsible AI acquisition process in Ireland has yet to move to enterprise-wide adoption with the preparation for full-scale governance and regulation.” “What organizations struggle most with is translating high-level responsible principles into repeatable and scalable operational practices. Even when executives recognize the importance of responsible AI, resource constraints, unclear ownership, and inadequate tools slow progress.”
The report found that Irish organizations are less likely than their US counterparts to have reached a stage where responsible AI is embedded in leadership and consistently implemented throughout their operating models.
The study also found that this is an area where Irish organizations are lagging in preparing for EU AI legislation. The Act regulates AI systems according to the risks they pose, and Irish companies using AI, even internally, must comply with the Act. The AI law will be implemented in stages, with full implementation by 2027, and penalties, including fines, will be imposed for non-compliance. Just 14% of Irish respondents admitted that their organization was fully prepared to comply with EU AI law. Another 70% said they were partially ready. The main challenges faced by Irish organizations when preparing to comply with EU AI law were limited in-house expertise (53%), budget/resource constraints (37%) and lack of clarity on EU AI law requirements (30%).
Mr Lee said that although there were challenges, attitudes towards responsible AI were changing among Irish businesses. “Leading companies are increasingly recognizing responsible AI as a strategic business enabler that delivers tangible value through enhanced trust, reduced regulatory risk, improved customer experience, and increased return on AI investment,” he said.
The overall survey was conducted among Irish business leaders and the results were compared to a previous similar PwC US survey. Just 16% of respondents in Ireland rated their organization as “very effective” in terms of standards for AI development and deployment, compared to 52% in the US. Other areas where there are significant gaps in AI governance implementation include communicating responsible AI priorities (Ireland: 21%; US: 52%). Employee training (Ireland: 14%, US: 49%), applying a risk-based approach to AI governance (Ireland: 33%, US: 47%), and having clear roles and responsibilities (Ireland: 28%, US: 52%).
Keith Power, Data and AI Partner at PwC Ireland, said: “This research highlights a clear next step for Irish organizations: a move towards responsible scale-up and enterprise-wide adoption of AI, supported by targeted investment in skills, governance capabilities and scalable execution models to meet regulatory demands and rapidly evolving AI technologies.”
