Zero capital, 116 projects: AI spots the devils in the details of China corruption case

Disciplinary staff in Wuhan, provincial capital of Hubei in central China, were going about their routine workday when a red alert landed. It was from a large data model – flagging a possible case of corruption.

“A company took on more than 10 projects this year. Its registered capital reduced from 50 million yuan (US$7.5 million) to 10 million, yet it’s paid-in capital is still zero,” the alert said.

“Moreover, over 80 per cent of the projects it took were [residential] communities from previous collaborations.”

The staff looked into the company – a construction firm in Wuhan – and found that it had taken on 116 repair projects in 84 residential compounds over the past five years, all with seemingly legitimate paperwork.

However, data analysis revealed it had few staff and no construction crew. Nor was there any trace that the repair works were ever carried out, according to the Communist Party’s top anti-corruption watchdog, the Central Commission for Discipline Inspection (CCDI).

The Wuhan disciplinary authorities launched an investigation to see if there was indeed any corruption involved and discovered that it was a shell company that some officials were using to misappropriate repair funds, the watchdog said.

The case was one of a dozen highlighted on the CCDI’s website this month, as authorities try out new ways of rooting out corruption, especially activities that are difficult to detect.

Party leaders and researchers in recent years have emphasised the role of big data in the anti-corruption fight. President Xi Jinping echoed the point in January when he called it an innovative tool for corruption detection.

Devendra Kumar, an associate fellow at the Centre of Excellence for Himalayan Studies at Shiv Nadar University, said CCDI had probably been experimenting with AI for some time.

“My sense is that this method has been experimented with and used by several branches of CCDI in the last two years and by now, the CCDI already has enough experience to use it at large scale,” Kumar said.

Meanwhile, a landmark document adopted at last month’s fifth plenum of the 20th Central Committee said “party building” would be stepped up across the tech sector nationwide and that digital tools would be deployed to fight corruption.

“We will promote the building of a digital system for discipline inspection and supervision and a big data platform for overseeing the exercise of power in the public sector,” the document said.

“All this will give us sharper eyes in combating corruption,” it added, while calling for continued monitoring of corruption-related risks and hidden dangers.

Citing the Wuhan case, the CCDI said that repair funds – paid for by homeowners and managed by the government – had traditionally been difficult to oversee because the money could be spent across several budget categories, making irregularities easy to hide.

03:00

China’s fifth plenum focuses on discipline and loyalty to Xi

The CCDI statement said the Wuhan government used artificial intelligence to monitor core areas – including the compound’s repair logs, construction staff changes and government operation – for anomalies that could expose potential issues.

AI has also been used in overseas cases. In January, state broadcaster CCTV reported that big data dug up the corruption case of Li Yong, former deputy party secretary and general manager of China National Offshore Oil Corporation.

Li disguised bribe-taking as project cooperation and hid the money in overseas accounts, making the funds difficult to trace and evidence hard to uncover. But big data models to analyse projects and business relationships helped the authorities establish that some businessmen had been especially close to him.

Among the projects he approved, one company was frequently the sole bidder, a pattern the authorities found suspicious. The investigators dug into the case and found evidence of corruption, the report said.

According to Kumar, it is only natural that the party’s disciplinary system will come up with such measures as the nature of corruption changes – for example, with the use of cryptocurrencies to hide illicit transactions.

“The disciplinary institutions are trying to catch up with changing, newer methods of corruption that cadres use to evade being caught than preemptively bundling such measures,” he noted.