The talent gap in network security is as high as 1.4 million How to alleviate the gap? _ID

Original title: How to alleviate the network security talent gap of up to 1.4 million?

China News Service, Beijing, September 14 (Reporter Li Jinlei) The digital economy is in full swing and AI is accelerating in people’s lives. But technology is a double-edged sword: it can be used for good or bad.

The security risks of AI cannot be ignored

Imagine these scenarios: the autopilot uses the lens and lidar to identify the target, but if a road sign is originally a left turn, the hacker deliberately uses artificial intelligence technology to add interference information so that the Recognition system mistakenly recognizes it as a straight road sign. It will become very dangerous.

For another example, face-changing AI technology is good for entertainment software, but will be problematic if used to create fake videos. If someone replaces the offender in the video with the face of a well-known celebrity, it will be misleading.

For another example, after using artificial intelligence technology to add noise to an image, you can trick the artificial intelligence system into making wrong decisions. For example, if a seller posts a pornographic image and uses technology to add interference, it can cause the AI ​​to misidentify it as clothing.

Nowadays, with the emergence of application scenarios such as facial recognition, motion recognition, false recognition and false image recognition, the results of the combination of artificial intelligence and security have become more fruitful, but the problems security facing artificial intelligence have gradually emerged.

Especially this year, the concept of “new infrastructure” has received increasing attention, including the new technical infrastructure represented by artificial intelligence, cloud computing and blockchain. The construction of a digital society around “new infrastructures” is accelerating. Under the digital infrastructure, the pressure on prevention and control of network security has increased dramatically.

Talent shortage in the cybersecurity industry may reach 1.4 million

However, security talent remains a scarce resource. For the frontier field of AI security, talents are even scarcer.

Alibaba Security Turing Lab senior algorithm expert Jiao Ao told a Chinanews.com reporter that statistics show that by 2020, the talent gap in network security will reach 1.4 million.

“Why is there such a shortage of talent? Because these people don’t understand safety, how to do it, and they lack opportunities and practical scenarios to hone their skills.” From this perspective, in addition to allowing more people to understand security and artificial intelligence, they should also be able to improve their security technology through specific scenarios.

In fact, the cybersecurity talent shortage is inseparable from the fact that the offensive and defensive experience is out of context and lacks systematization. Many industry experts believe that today’s security industry is in strong demand for talent, and the training cycle of an outstanding security professional is very long and requires years of systematic experience in rich real-world business scenarios.

Combining industry and education to create the world’s leading AI challenge to empower talent

It is an effective way to solve the talent gap and resist the security risks that AI will face in the future by holding various competitions and competitions, providing real combat scenarios and promoting learning through competitions, thus discovering and training. more security talents.

For example, Alibaba Security and Tsinghua University co-organized the Security AI Challenger program, with adverse samples as the core, providing scenarios, technology, data, computing power and other support, providing a testing ground for the security of digital infrastructure, in a real environment with high difficulty. Enhance technology and cultivate security infrastructure talents with real security capabilities.

The Security AI Challenger program is reported to have conducted the first phase of the face confrontation recognition competition, the second phase of the ImageNet image confrontation competition, the third phase of the abusive scene text confrontation competition and the fourth phase of general target detection confrontational attacks, attracting 200 worldwide. More than 5000 teams from more than 100 companies from many universities participated. The fifth contest against the falsification of images of documents and documents began on September 7.

“The questions we set are all based on the real system. They are all part of the real system. We want participants to have the opportunity to attack similar real systems, and their behavior will not be illegal, so he has the opportunity to increase his experience and hone your skills, “said Juao.

Alibaba’s senior security expert Hunter said that the most important feature of the security AI challenge program is the series of long-term competitions with AI, which helps to stimulate the creativity of the challenge in the field of AI of security and helps solve real business problems. Strengthen, fill, and alleviate the new talent gap in infrastructure security to the greatest extent possible. (Finish)Return to Sohu to see more