A substantial majority of Irish compliance professionals believe that criminals are moving faster than companies and regulators in exploiting artificial intelligence, according to a new survey by the Compliance Institute.
The survey found that 77% of respondents think businesses and regulators are lagging behind, while an additional 20% said criminals are “somewhat” ahead at the moment. Michael Kavanagh, CEO of the Compliance Institute, said cybercrime has become the most prevalent crime in Ireland and highlighted how AI is being used to create deepfake audio and video, generate forged documents, and even build detailed psychological profiles of individuals by scraping social media accounts.
“Criminals are now capable of producing convincing IDs, payslips, and other documentation,” Kavanagh said. “All of this can be leveraged across multiple sectors, and the speed at which these technologies are evolving is alarming.”
The survey also explored perceptions of responsibility for errors linked to AI. Nearly a third of professionals said companies deploying AI should bear ultimate responsibility, compared with 11% who attributed responsibility to developers and just 3% to regulators. Over half of respondents felt accountability should be shared among all stakeholders.
Kavanagh stressed that companies need to take the lead in ensuring AI is used responsibly. “The onus is on the company as opposed to the vendor of the AI system or the regulator to put the processes in place,” he said. “They need procedures, they need processes, and they need to continually update them.”
While there has not yet been a legal test in Ireland determining where liability lies for AI-related errors, Kavanagh said firms are generally expected to stand by their own systems and practices. He emphasised the importance of human oversight and proactive staff training.
“You need to constantly update your training in this area and continually stress-test the AI systems in place,” he said. “Human intervention is essential. Machines have not taken over yet.”
Failure to maintain oversight, he warned, can expose firms to serious reputational and financial risks. Over-reliance on AI without proper controls could lead to regulatory fines under the EU AI Act, as well as potential legal liability and damage to public trust.
“The reputational risk of companies is enormous, and the legal liability is real,” Kavanagh said. “Regulatory fines are coming, and firms must ensure they are prepared.”
The survey’s findings highlight the growing tension between rapid technological advancement and the frameworks designed to manage it, signalling an urgent need for businesses to strengthen AI governance while keeping human judgment at the centre of decision-making.
