China develops the world's first artificial intelligence 'prosecutor'

Chinese researchers say they have developed the world's first machine using artificial intelligence (AI) technology that can prosecute and prosecute criminals. Crafted and tested by the Shanghai Pudong People's Procuratorate, the AI ​​'prosecutor' is said to be able to bring charges with over 97% accuracy based on verbal descriptions of the case. .

According to Professor Shi Yong, Director of the Data and Knowledge Management Laboratory of the Chinese Academy of Sciences, the technology can help reduce the daily workload of prosecutors, allowing them to focus on more difficult tasks.

Prosecutors were early to use AI in law enforcement when the technology was first discovered in China in 2016. Many agencies in the country are using an AI tool called the System. System 206, can analyze the authenticity of evidence, assess the conditions of arrest, and how dangerous the suspect is to the public.

However, all existing AI tools have a very limited role because 'they cannot participate in the decision-making process of charging or recommending sentences,' Professor Shi Yong and colleagues said. . Making such decisions would require a machine that could identify and rule out any content of the case file without omitting important information. In addition, the AI ​​machine also needs to be able to convert complex, different types of human language into a standard mathematical or geometric format that the computer can perceive.

China's Internet companies have developed many modern tools that can process natural language, but these operations often require working on mainframes that prosecutors don't have access to .

The AI ​​'prosecutor' developed by Professor Shi Yong's team can work on regular desktop computers. For each suspect, the system will make accusations based on 1,000 characteristics obtained from human-generated case description text, even if it is very small or abstract data. The 206 system then evaluates these evidences.

The new artificial intelligence machine has been 'trained' for use in more than 17,000 different cases between 2015 and 2020. So far, it has been able to identify and convict the eight most common crimes in Upper Sea. These are credit card fraud, organizing gambling activities, dangerous driving, intentionally causing injury, obstructing law enforcement, theft, fraud, quarreling and disorderly conduct.

According to the researchers, the AI ​​'prosecutor' will soon become more effective with upgrades. It will also be able to recognize less common crimes and bring more charges against a single suspect.

However, an unnamed prosecutor in the southern Chinese city of Guangzhou, said he had some concerns about the use of AI when prosecuting criminals.

'97 % accuracy may be high from a technological point of view, but there will always be the possibility of error. And who will be responsible when something goes wrong? Prosecutor, machine or algorithm programmer?', he shared.

In addition, the inclusion of AI directly in the prosecution process may also affect the prosecutor's operating power. In fact, most prosecutors don't want computer scientists to 'get their hands on' legal decisions. Another problem is that the AI ​​prosecutor can press charges based solely on its previous experience. Specifically, the machine cannot anticipate the public's reaction to an incident in a changing social environment.

"AI can help detect errors, but it cannot replace human decision-making," said the Guangzhou-based prosecutor.

Many Chinese courts have been using AI to help judges process case files and make decisions such as accepting or rejecting appeals. Most prisons in China have also adopted AI technology to monitor inmates' physical and mental condition, with the goal of reducing violence in prisons.

Not only in China, the application of AI technology in law enforcement is increasingly popular around the world. Some German prosecutors have used AI technology for image recognition and digital forensics to speed up the speed and accuracy of case processing.

In fact, AI technology can have incalculable consequences for humanity, with the risk of replacing humans in the future. Before that, in 2017, the story of Sophia (operated by an AI system) became the first robot to be granted legal citizenship (even more rights than women in this country). stir up public opinion. The humanoid robot Sophia, who is a Saudi national, has made some controversial statements, but most recently, a statement from Robot Sophia left the world speechless: I want to have it. a robot child and start a family.

In a conversation with David Hanson, the creator of this robot, said that Sophia will destroy humans. Of course, that has caused anxiety, because since the early days of robots and AI technology, people have feared that they will rebel against humanity.

The robot has also made controversial appearances as lecturers on science and technology topics, appearing alongside celebrities in video conferences, TV shows and universities. study around the world. In early 2021, Hanson Robotics announced that it would start producing hundreds of robots using AI technology like Sophia.