Experts say AI can streamline the detection of corruption, or conceal it, at forum in Seoul

Legal experts and officials warned at an international anticorruption forum in Seoul on Friday that AI can pre-emptively prevent wrongdoings but also enables “concealing corruption" or avoiding responsibilities.

"AI has two faces when it comes to anticorruption policy," Choi Yong-jeon, president of the Korean Anti-Corruption Law Association, said during an international anticorruption forum organized by the Anti-Corruption and Civil Rights Commission (ACRC) at Lotte Hotel Seoul in central Seoul. "AI enables preventive anticorruption. Traditionally, anticorruption efforts have focused on detecting and punishing corruption after it has occurred. However, AI can analyze procurement, public finance and audit data to identify warning signs in advance such as conflicts of interest, repeated contract awards to the same bidder and abnormal pricing patterns."

Choi also stressed the challenges AI can present for anticorruption efforts in the forum, which shared practices to inspire global governance reform and highlighted major outcomes from the partnership between the ACRC and the United Nations Development Programme Seoul Policy Centre, focusing on tackling corruption worldwide. This year’s event took place on the 10th anniversary of the collaboration.

"AI is not intended to replace the judgment of auditors or investigators," Choi said. "Rather, it serves as a supporting tool that provides early warning signs of potential risks. Its effectiveness depends on several factors, including access to sufficient data, minimizing both false alarms and overlooked cases and ensuring a transparent model."

"AI can also become a new structural risk. If algorithms are opaque and accountability is unclear, AI may cease to be a tool for exposing corruption and instead become a mechanism for concealing it."

If, for example, AI selects the winner of a bid without a valid explanation, no one can verify whether the decision was impartial.

Choi, however, highlighted that AI can serve as an effective tool to reduce human efforts in uncovering corruption, while stressing the need for measures to govern AI.

"AI can analyze large volumes of administrative data to identify abnormal patterns that are difficult for humans to detect," Choi said. "Whether AI becomes a tool for corruption or a foundation for public integrity depends not on the technology itself, but on the legal and institutional frameworks that govern it. In the age of AI, combating corruption is not simply a matter of adopting new technology."

Other officials at Friday’s session, such as Hong Jeong-ki, director of the digital and AI procurement management division of the Public Procurement Service of Korea, and Azzaya Ganbold, head of the National Anti-Corruption Strategy Implementation Department of Mongolia, shared their views on harnessing AI and digital technologies for anticorruption efforts.

The officials identified other priorities that the public and private sectors must be aware of in accordance with the use of AI.

The forum was a two-day event, attended by current ACRC Chairperson Jung Il-yeon and former ACRC Chairperson Kim Young-ran on Thursday. During Kim's tenure at the ACRC, she proposed a landmark anticorruption law aimed at reducing bribery and improper influence known as the Kim Young-ran Act, which prohibits public officials, teachers or employees of public institutions from accepting expensive gifts or money that could influence their decisions.

BY PAIK JI-HWAN [[email protected]]