- Design transparency, accountability, controllability before efficiency
Adoption of new technologies such as Artificial Intelligence (AI) tools alone will not create legitimacy and reduce corruption without effective legislation and democratic safeguards implemented in parallel, experts warn.
Daejin University Department of Law and Public Service Professor and Korea Anti-Corruption Law Association President Choi Yong-jeon, speaking at the International Anti-Corruption Forum in South Korea on Friday (3), pointed out that technology adaptation without proper constitutional legislation could allow for new forms of corruption to grow and act as a smokescreen to others.
He advocated that democratic legitimacy and accountability must come before technical accuracy.
Commenting on the use of technology to streamline governance and improve transparency, Choi told the gathering organised by the United Nations Development Programme (UNDP) and South Korea’s Anti-Corruption and Civil Rights Commission (ACRC) that AI had a ‘dual nature’ which could be seen as both an anti-corruption tool and as an emerging corruption risk. He highlighted the prospective risk that AI tools, if not properly regulated and used with an effective transparency structure, could be misused.
Discussing the use of new technologies for anti-corruption in a panel discussion of experts, Choi explained that AI could be used to improve corruption prevention, with it used for post hoc detection and risk management, and to analyse public procurement, fiscal, and audit data.
AI can be used to detect collusion, conflicts of interest, and repeat-award patterns early, while AI tools can also be used to allocate audit and investigative resources strategically for effective investigations. AI can help detect abnormal patterns at an early stage and direct audit resources towards high-risk areas.
However, Choi warned that the adoption of AI with effective legislation and institutional oversight could lead to AI becoming a new power structure that concealed corruption and made it harder to detect.
He pointed out that anti-corruption systems could become channels for information monopoly/surveillance and evidence manipulation if safeguards and control management were not effectively implemented. “Rather than presuming AI to be a ‘fair tool,’ we should treat it as a normative object that reshapes power and accountability,” Choi opined.
UNDP Global Programme Adviser on Anti-Corruption Anga Timilsina told the gathering that there was a need for stronger synergies between national digital or AI strategies and national anti-corruption strategies, with improved awareness and understanding of AI and related technologies in the Global South to effectively employ AI tools in governance.
He highlighted that sustained investment in AI tools and technology platforms used for digital governance was required and stressed that there was a need to clearly define objectives and scope for AI initiatives and human verification, which should be kept as a safeguard.