During the Double 11 period, many people downloaded the shopping app to buy it. Who knows that the downloaded shopping app not only did not wait to buy the goods, but also lost a lot of money for no reason. According to relevant data released by the 360 security brain, in the month before this year's Double 11, the number of active fake counterfeit mainstream shopping apps was close to 4,000, and these apps have covered more than 300,000 mobile devices. According to 360, the source of these APPs is unknown. There may be shopping risks such as stealing user account passwords, such as privacy and fraudulent money, and it may also hide the virus.
With regard to the frequent occurrence of citizen personal information cases in recent years, data security experts have pointed out that the main threat to data security and privacy protection is still the increasingly rampant black-gray gang behavior.
Black ash gangs become the main threat to data protection
"The leaked data will mostly be used by black and gray gangs for fraudulent acts such as fraud, which will cause serious harm to individuals, society and the country." At the 2018 (3rd) Data Security and Privacy Protection Conference held recently, it was black. The gray industry has extensively discussed the threats posed by personal data protection.
According to the Nanfang Daily reporter, whether it was the disclosure of 3 billion user information disclosures involving 96 Internet companies in August this year, the first domestic empty number SMS hijacking case involving millions of fake online accounts, and more than 500 million in China. The user's privacy data disclosure case, or the outbreak of the 100 million-level information and data disclosure case involving the virus intrusion in the 10 real estate system in Shandong Province in September, there are black gray gangs behind the scenes.
The "Summary Analysis of Sensitive Information Disclosure Cases at Home and Abroad" pointed out that the Internet industry, government agencies and industries such as finance, education and medical care are the hardest hit areas for sensitive data leakage. Among them, more than 70% of personal information leaks are caused by hacking and other means, and another 20% are caused by non-technical means such as inserting or buying "inner ghosts", and there is still a 7.87% leak path.
In March last year, the Ministry of Public Security launched a special campaign to crack down on hacker attacks and cyber violations of citizens' personal information crimes. In just 4 months, it detected more than 1,800 related cases, arrested more than 4,800 criminal suspects, and seized more than 500 personal information of various citizens. Billion. Third-party enterprises and organizations without security protection are the main targets of black-and-grey gang attacks. These enterprises usually have a large amount of data, but often lack sufficient data security awareness, which makes the overall data security protection of the Internet industry become stronger. It is imminent.
"Data poisoning" makes artificial intelligence go wrong
From the era of interoperability to the era of artificial intelligence today, various data, including personal information, have become the most valuable asset. But data, as a development factor that many areas of the Internet industry rely on today, are also threatened by "pollution."
"Artificial intelligence has given us empowerment, and its own security problems have been neglected. At present, artificial intelligence is essentially data intelligence, and it is necessary to prevent data poisoning." Zhou Hongyi, chairman of 360, said that when talking about future network security, The whole industry should take a cautious attitude towards the development of artificial intelligence, especially in terms of security.
It is understood that some experts have predicted that there will be tens of billions of smart devices connected to the Internet by 2030. With the development of artificial intelligence technology, software defines everything, everything needs to be interconnected, and network security will also pose a huge threat.
Zhou Hongyi said that artificial intelligence is trained by big data, and the training data can be polluted, also called "data poisoning." For example, by inputting a “pollution” training sample to a smart car, the “no-pass” traffic sign can be identified as “passable”, resulting in traffic accidents and casualties. At the same time, sensors that artificially identify data can also be disturbed. The same is a smart car, with a laser pointer swaying against the laser radar of the driverless car, the car may stop in an instant and be chased by the rear car. In addition, Zhou Hongyi also said that the complex computational logic and decision-making process within artificial intelligence is still opaque and can be deceived or bypassed. In this way, the home smart sound can even become a bug. Zhou Hongyi believes that the future world is a world that is swallowed up by software and full of loopholes. There is no impregnable line of defense. For network security, the power of new technologies such as artificial intelligence, big data and cloud computing is also needed. (Reporter Ye Dan)
0 comments:
Post a Comment