[Abstract] The use of artificial intelligence technology to detect fraud and crime is a process of mutual wits, and we should quickly determine how to use artificial intelligence to prevent the basic principles of fraud and crime. Tencent's artificial intelligence is a technology that will become mainstream in the future. It can appear anywhere in homes and cars. Although artificial intelligence technology is not a panacea, applications that use artificial intelligence technology to help identify fraud and lies can still occur in the future. The use of artificial intelligence to detect fraud and crime is a process of mutual ambition, and we should determine as quickly as possible how artificial intelligence can prevent fraud and crime. At present, there are mainly three ways to use artificial intelligence to fight fraud, and they play an important role in the development of artificial intelligence. they are, respectively: 1. List of rules and reputations; 2, machine learning methods; 3, non-directed machine learning behavior; Rules and reputation list Rules and reputation Recognizing fraud in many organizations is an effective behavior. It is very similar to the "expert system", which was first introduced in the field of artificial intelligence in 1970. The expert system refers to the combination of computer programs and expert experience in different fields. The entire system is easy to start and run, and is based on human comprehension, but it is also limited to areas such as high manual labor. "Rule" is a behavior that detects fraudulent accounts with human logic and statements. For example, an organization can set a rule that when an account costs more than $1,000 to buy a project, the project will be blocked within 24 hours. Machine learning method Machine learning is a branch of artificial intelligence technology that can prevent the system from solving problems in an overly rigid manner. Researchers want the machine to learn that data instead of looking for an expert system through computer coding. Machine learning technology has made great progress since the 1990s, and it has been widely used in anti-fraud activities since the 21st century. Unsupervised machine learning behavior Unguided machine learning can be used in fewer areas, including fraud detection and prediction of the results of multi-layer tests. Unguided machine learning behavior is difficult to crack internally because it needs to deal with billions of simultaneous possibilities without the need for special guidance mechanisms. Some companies have made great progress in this area. Source:venturebeat