In China, it’s possible to do almost anything with WeChat: call a taxi, pay bills—even get one’s day in court.
Going to trial with WeChat-based courtroom mini-app Weisu (微诉) looks a lot like a Skype meeting: courtroom participants can join from the comfort of their couch. The mini program—which functions as an independent app but can be launched directly from the social platform—can verify their ID, submit court files, and have their testimonies transcribed by WeChat’s voice-to-text technology.
Weisu—developed by Chinese big data and AI company Gridsum—is part of a larger initiative to digitalize Chinese courts making them faster and more efficient—as well as collect big data. For now, the software is largely used in intellectual property litigation.
One well-known example of the effort is China’s first cyber court launched last year in the Chinese city of Hangzhou—home of e-commerce giant Alibaba. The court was set up to handle the swelling number of online disputes including e-commerce complaints, online lending disputes, and copyright infringements.
(If you don’t see a video above, try watching on QQ video instead.)
Aside from remote trials, Chinese courts started live streaming some of its trials in 2016 as a part of an effort to shed some light on Chinese judicial system which has often been criticized as opaque. China has a 99% conviction rate, more than any other country in the world.
“Streaming of cases improves transparency but for lower courts, streaming is selective,” said Susan Finder, a China-focused legal scholar that has been monitoring the Supreme People’s Court for more than 25 years. “How many people pay attention to the case streaming is also a question.”
Many other technologies are being integrated into the judicial process. In March, a court in Beijing trialed a VR visualization of a crime scene. Hoping to mitigate the lack of lawyers for enterprises, e-commerce platform JD recently presented its voice recognition-powered AI legal bot Fadongdong (法咚咚) which is also available as a WeChat mini program. Alibaba Cloud’s ET Brain has equipped 6000 courtrooms with its AI speech transcription. The legaltech industry is rising: China filed 34% of legaltech patents globally in 2016, second only to the US, and the demand is likely to grow.
Among these technologies, artificial intelligence is grabbing the most attention but it is also one of the more controversial ones. Gridsum, for example, is providing judges AI-powered suggestions on how to handle their cases with the help of a legal search engine based on data from China’s court database opened in 2014.
“Our core competence is to generate data based on text mining, turn that data into knowledge, and then make it available to the court, legal practitioners, and then to judges and court presidents and others,” Du Feng, deputy general manager at Gridsum Legal Big Data Division, told TechNode.
However, even a machine would find it hard to sort through a legal system, and not just China’s. As Gridsum’s Du explains, an AI machine can quickly grasp how to play a game of Chinese chess or Go like the example of Google’s AlphaGo, but China is too big; each province has its rules and guidelines on verdicts with great development gaps among different provinces.
Geographic differences aside, laws, regulations, and the judicial system as a whole is constantly changing and upgrading. This makes providing AI assistance to judges a complex endeavor, according to Du. “Our product has to be precise,” he says.
Unlike some countries, China has not yet decided to hand over more important decision making to AI, according to Finder. Police and judicial bodies in the US, UK, and soon in Switzerland are using risk assessment algorithms for decisions about pretrial release and parole including software such as COMPAS, HART, and ROS.
“The use of an algorithm in crime prevention should help to eliminate uncertainties when authorities and judges have to make difficult decisions,” said Ioannis Martinis, a lawyer for the Swiss legal protection insurance Coop Rechtsschutz where he leads artificial intelligence projects. “But the use of such algorithms seems and feels like delegating human responsibility to the machine—that is always somewhat questionable.”
Both in the UK and US, the use of AI systems in making legal decisions has come under attention for racial bias and bias toward people from poorer areas, urging developers to be aware of the possible prejudices that get built into algorithms.
Another risk of using AI decision making in court is the risk of reducing or even completely abandoning human control due to cost and/or time pressure, according to Martinis. Considering China’s lack of legal experts, this scenario is not hard to imagine.
However, interest in solutions such as these is growing in China. The AI “black box” was one of the main topics the AI and Law forum at this year’s World Artificial Intelligence Conference in Shanghai which brought speakers from China’s judicial system, law research institutes, and legaltech startups.
However, most experts at the forum agreed that what China needs is practical solutions and merging the convenience of WeChat with courtrooms is one of them. As CEO of PowerLaw AI Tun Cunchao said during the forum, AI and law need to focus on solving specific tasks rather than “fantasizing about an omniscient and omnipotent legal AI solution that can be used universally.”