As corruption schemes become increasingly sophisticated and digital in nature, anti-corruption experts say artificial intelligence (AI) is emerging as a valuable asset in strengthening investigations. However, they emphasise that the technology should complement, rather than replace, the judgement and decision-making of investigators.
The role of AI in tackling corruption was a key focus at the Asia Regional Expert Meeting on Measuring Effectiveness of Anti-Corruption Agencies, which concluded yesterday.
Participants highlighted AI’s ability to process and analyse massive volumes of information within seconds, enabling investigators to identify suspicious patterns, financial irregularities, and potential risks far more efficiently than traditional methods.
According to experts, AI can support investigations by detecting anomalies in government procurement processes, cross-checking public officials’ asset declarations against financial records, identifying conflicts of interest, and uncovering concealed connections between officials, contractors, and suppliers.
“Artificial intelligence can process enormous datasets in a very short time, significantly reducing the workload of investigators,” said Drago Kos, Dean of the International Anti-Corruption Academy. “It helps detect risky behaviour, accelerates the search for evidence, and strengthens the credibility of investigative findings. Unlike humans, it does not become fatigued or make routine errors.”
Beyond corruption investigations, AI is increasingly being deployed worldwide to enhance anti-money laundering efforts, improve auditing processes, analyse legal documents, monitor public infrastructure projects, and strengthen digital platforms used for reporting corruption.
For Bhutan, where corruption and fraud cases are becoming more data-intensive and financially complex, officials from the Anti-Corruption Commission (ACC) said the discussions were particularly timely.
The Commission has already begun integrating AI-assisted technologies into its investigative work. These systems are capable of analysing bank statements, converting unstructured financial information into usable data, retrieving records from civil registration databases, and mapping relationships between individuals to expose hidden networks that may otherwise go undetected.
Despite these advances, experts cautioned against overreliance on technology. While AI can improve efficiency and provide valuable analytical support, it cannot establish criminal liability or replace professional judgement.
Kos stressed that investigators must remain responsible for every stage of the investigative process, using AI solely as a decision-support tool rather than an independent investigator.
“AI will not investigate cases on behalf of investigators,” he said. “Its purpose is to assist. Investigators must remain in full control of how the technology operates and how its outputs are interpreted. The final judgement must always rest with the investigator.”
As Bhutan continues to modernise its anti-corruption efforts, experts noted that the successful adoption of AI will depend not only on technological capability but also on maintaining robust ethical standards, effective human oversight, and strong institutional accountability.
