马来西亚 MACC 扩大 AI 应用，推进情报主导型调查

Malaysia’s anti-corruption agency is expanding its use of AI, data analytics, and digital systems as part of efforts to strengthen investigations and identify corruption risks earlier.

Malaysian Anti-Corruption Commission (MACC) chief commissioner Datuk Seri Abd Halim Aman said technology adoption is one of five priorities guiding the agency, alongside integrity, fair enforcement, proactive prevention, and public trust.

The technology is being introduced alongside MACC’s wider use of Intelligence-Based Investigation, or IBI, which reduces its reliance on complaints or formal reports to initiate investigations. Abd Halim said the commission wants to detect and uncover corruption earlier and act before cases escalate.

The approach predates MACC’s latest AI push. Former chief commissioner Tan Sri Azam Baki said in 2025 that MACC had adopted IBI techniques, particularly for complex and high-profile corruption cases, with the process including the identification of potential subjects and efforts to recover misappropriated assets.

From reports to intelligence-led investigations

Abd Halim said combining data analytics, AI, and digital systems allows investigators to process large volumes of information more quickly. The tools are being used to identify anomalies, trace financial flows, and detect connections that can be difficult to uncover using conventional investigative methods.

MACC said IBI allows investigators to move beyond waiting for reports, while MACA training material describes the method as involving the systematic collection and analysis of intelligence to identify patterns, anomalies, and corruption risks.

MACC had already outlined plans to expand its use of these technologies earlier this year. In July, Abd Halim said digital technology, data analytics, and AI would be used more widely across intelligence, investigations, digital forensics, prosecution, and organisational management.

Abd Halim said investigations into 1MDB and SRC International expanded MACC’s expertise in financial forensics, cross-border investigations, and asset recovery.

MACC has identified government procurement, enforcement agencies, public funds and special allocations, and government revenue as four areas of operational focus in 2026.

Between January and August 31, the commission received 5,239 reports and opened 842 investigation papers, leading to 803 arrests. The cases included 352 involving the receipt of bribes, 278 involving false claims, and 78 involving the offering of bribes.

Between January and July 31, MACC also recorded RM55.3 million through forfeitures, compounds, settlements, asset freezes, and seizures.

Data access and governance behind AI investigations

Data quality is also part of Malaysia’s wider framework for public-sector AI adoption. Government guidelines identify it as one of the requirements agencies need to address when deploying AI systems.

The same guidelines identify digital infrastructure, talent development, leadership support, and change management as other requirements for adoption. They do not state that MACC has problems with its data.

Malaysia has also established a broader framework for sharing data between federal agencies. The Data Sharing Act 2025, or Act 864, came into force on April 28, 2025, and provides legal authority for public-sector agencies to exchange data under defined governance requirements.

Under the law, agencies requesting data must state what information is required, why it is needed, which agencies are supplying and receiving it, and how it will be stored, used, archived, and destroyed. Third-party technology providers handling shared data require prior written consent from the supplying agency and remain subject to security and confidentiality requirements.

The government also operates the Malaysian Government Central Data Exchange, or MyGDX, which allows agencies to exchange information through application programming interfaces. Data is referenced directly from its original authoritative source, while the platform’s catalogue includes datasets covering procurement, companies, population, land, employment, vehicles, and public administration.

MyGDX 2.0 uses identity verification, digital certificates, and encrypted connections as part of its access and security controls. There is no public evidence that MACC’s AI or IBI systems are directly connected to MyGDX.

Malaysia’s public-sector AI rules also classify law-enforcement applications as high risk. JDN’s Public Sector AI Adoption Guidelines divide applications into prohibited, high, limited, and minimal-risk categories, with law enforcement among the uses subject to stricter controls.

The framework requires agencies to address areas including data privacy, security, transparency, accountability, fairness, and system reliability. It does not provide details on the specific AI models or analytical systems used by MACC.

Building skills and keeping humans in the loop

MACC is also developing staff skills to support greater use of AI and analytics. Its Malaysian Anti-Corruption Academy held a five-day advanced AI programme in July this year for 13 officers, covering areas including information and document management, data analysis, and report preparation.

The training complements MACC’s existing programmes on intelligence-based investigations. MACA material shows that IBI training covers intelligence assessment, analysis of relevant data, and the identification of information that can be used to develop investigative cases.

MACA has separately provided training in AI-assisted data analysis and intelligence-based investigation methods.

Talent development is also included in Malaysia’s wider public-sector AI framework. JDN guidelines list it alongside data quality, digital infrastructure, leadership support, and change management as areas agencies should address when implementing AI systems.

Government statements in 2026 also said AI used in higher-impact public-sector functions should support analysis, while final decisions remain with authorised officials.

MACC has not publicly said that AI will independently determine whether an individual becomes the subject of an investigation. Its public description instead presents AI and analytics as tools used by investigators to process information, identify anomalies, trace financial flows, and detect relationships within available data.

MACA’s IBI training similarly focuses on assessing intelligence, analysing data, and identifying information that can support the development of a case.

Abd Halim has placed technology adoption alongside transparency and public trust among MACC’s five priorities. He said the commission’s use of technology is intended to improve operational efficiency and transparency in investigations.

(Photo by Ervince Kang)

See also: MIT Transit Lab secures $2.1M from Google for AI transit platform

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.