It is particularly important to control the source of information leakage. Industry calls for legislation before algorithm.
■IT Times reporter Zhang Weiwei
In the last two years, Tencent has raced around in the security field. Some insiders joked that Tencent's security laboratory is enough to occupy half of the whole industry.
With the increasing strength in the security field, Tencent has made continuous attacks to create a "guardian plan" to crack down on online black production. In particular, on 20 17, the "Guardian Plan" moved the center of gravity of the network black production management forward and intensified the attack on the threat source.
Strengthen the attack on the source of black production
20 17 Tencent is behind several high-tech crimes in China.
In 20 17, a coding platform using artificial intelligence technology to crack Internet verification code appeared in China. According to the police, when stealing users' privacy, black gangs usually get users' personal accounts by crashing the background database of the Internet platform, but the dynamic verification code blocks the invasion. In order to break this line of defense, black gangs hired people to input matches, but last year, a black gang developed a machine that can quickly identify the verification code picture, encode it (enter the password), and refresh a large number of account verification codes in a short time. An artificial intelligence system can only be trained in one to two months, which is extremely harmful. After extensive promotion of various social platforms, this black gang quickly occupied 90% of the black market. In the first quarter of 20 17, the coding system cracked the verification code 25.9 billion times.
Li Xuyang, general manager of Tencent's security cloud department, said that from the current security situation, after the emergence of a new technology, black products are often the first to make profits and need stronger technical strength to attack from the source. Internet companies can do a lot for this. For example, the security team of Tencent's "Guardian Program" models the data of its social platform, locates it at different levels, and cooperates with the police to find the source of black production.
Similar cases include the "Trojan-based reading platform", which can break through the user's mobile phone, brush reading, praise and vote without the user's knowledge. 20 17 in may, Tencent security team assisted Jiangsu police to successfully detect the case of WeChat Trojan brushing official account traffic, and smashed a number of criminal gangs that developed and operated Trojan programs. "Through big data analysis and clustering, listing the core backbones in the black production industry chain is conducive to the public security organs to carry out offline strikes, so as to achieve a combination of attack and prevention, with good results." Zhu Jinsong, general manager of Tencent's security management department, revealed that Tencent has been exploring how Internet companies can give full play to their technical and data advantages in public security attack and defense since it established a public welfare platform against telecommunication network fraud on October 6th.
Hold the safety bottom line
At present, Tencent has seven laboratories, including Security Cohen Lab, Xuanwu Lab, Zhan Lu Lab, Security Genting Lab, Anti-fraud Lab, Anti-virus Lab and Mobile Security Lab, covering cloud computing security, communication anti-fraud, website anti-fraud and pseudo base station positioning. Tencent's goal is to create a new domestic security ecology, and it has made frequent moves in the field of social public security in the past two years. In 20 17, Tencent and the State Administration for Industry and Commerce established a monitoring and governance base for online pyramid selling, and set up joint security laboratories with Shanghai Anti-Telecommunication Network Fraud Center and shenzhen public Criminal Investigation Bureau respectively.
"After the 3Q War, Tencent invested a lot of financial resources in the security field. At present, Tencent can compete with 360 in the security field. " In the eyes of the industry, Tencent's investment in the security field is the most in BAT. Through cooperation with government departments, Tencent has exported the results of secure big data, which has both social benefits and brand value.
However, cracking down on the black production team cannot completely eliminate information leakage. Especially with the development of technology and industrial upgrading, it is necessary to put legislation and other work ahead.
In fact, the development of artificial intelligence is inseparable from the support of user data. For example, at present, portal websites use users' reading and browsing data and artificial intelligence algorithms to recommend news information to users individually, but how to define whether it is illegal for artificial intelligence to call user data? At present, there is no clear boundary in the industry. Zhu Jinsong said frankly that at present, the disclosure of citizens' personal information is very serious, and it is easy for black gangs to obtain users' portraits on the Internet, knowing users' names, ID numbers, children's reading information and so on. The solution to the problem lies not in destroying several black market gangs, but in the governance and control of the source. "Academia is discussing whether legislation can be put in front of the algorithm, control risks before the algorithm is actually applied, and keep the safety bottom line as much as possible. This can help China play a guarantee role in the global artificial intelligence competition. " Zhu Jinsong said.