“AI takes up the banner of cyber security. Start-up companies are the first to eat crabs with a valuation of 1.2 billion US dollars. Recently, cyber security has been at the forefront of the topic list.

“

Recently, network security has been at the forefront of the topic list.

In 2015, cybercrime caused US$3 billion in losses throughout the world. If the momentum continues, by 2021, this number will become 6 billion per year (including data loss and destruction, money loss, inefficiency, IP Theft, etc.).

In March 2018, the Internet Security Vulnerability Report issued by Symantec Cyber ​​Security stated that “Every year, not only the number of cyber attacks increases, but the targets of attacks are also more diverse. Attackers are looking for attack paths and cover up. I work harder on my whereabouts.”

It is always said that AI technology can be the protector of data security, and the British company Darktrace is a good example of it. Since its founding in 2013, Darktrace has developed rapidly. Now there are more than 700 employees worldwide, and the company is valued at US$1.25 billion. Although many contracts worth several million US dollars have been signed, the company refuses to disclose its revenue to the outside world. .

In the end, in this battle between righteousness and evil, those who win AI win the world

The company’s co-founder and CEO Nicole Eagan told CNBC, “We are now in a cyber arms race. The battlefield is the network system of each company. This is a war on algorithms.”

The principle of Darktrace’s AI technology is briefly described as follows: From the cloud or the company’s internal network, or even a dedicated factory, the company’s technology will protect the company’s internal personnel from using the network by preventing external attackers from finding a way to enter the company’s computer. The purpose of not being disturbed.

Complex technology actually escapes from the most common thing around us: the human immune system.

As we all know, the immune system can recognize what is “self” and what is “outsider”, and can also recognize viruses and threats that have never been exposed to. Darktrace’s system is also based on this principle. It deeply learns the habits of every computer user and every device in the company, and learns what “self” is from this process. This technology can distinguish peculiarities in real time, thereby preventing threats from entering.

Of course, the Dao is one foot high and the magic is high, and the attacker’s methods are getting higher and higher.

Because of this, Dartrace also presupposed the expectation that the attacker had actually entered the company’s network. “A software service function called SaaS will work’from the inside out’ to find out when the attacker came in, what information they got, and stop it before causing serious damage.” This is like our body. Immune system, sometimes we catch a cold, viruses do enter our body, but the immune system will find their disguise, and then deal with these uninvited guests as soon as possible.

As a result, Darktrace finds new threats every day.

For example, the entire company network is like a giant fish tank, and the protection mechanism is like a fishing net in a tank. Once, a customer was attacked. Darktrace caught the “fish that slipped through the net”, only to discover that it was the company’s internal personnel who added an IoT thermostat device but forgot to tell the management department.

However, a real threat is that the attacker sneaked in from this thermostat and wandered through the “rich man” database. And upload the data to the cloud. Darktrace discovered this anomaly because the general thermostat does not go to the “rich man” database.

On another occasion, an employee of a major bank had been using the company’s server to mine bitcoins, but this behavior was different from other data center servers in Darktrace’s eyes. After being discovered, the employee’s dream of making a fortune was shattered.

Darktrace is not the only player in this field.

“We see more and more companies interested in the application of machine learning, deep learning, and artificial intelligence in the field of cyber security.” The director of the University of Maryland Cyber ​​Security Graduate Program told CNBC. The university’s computer science professor and director of the Cyber ​​Security Center added, “There are already multiple companies using AI and machine learning to combat cybercrime. Although there is no accurate data, it has indeed increased in the past five years.”

Trustlook in California is an example. It was also founded in 2013 and raised more than 35 million U.S. dollars. The ending plan of this company mainly protects mobile devices, Internet devices, and the Internet of Things. Like other machine learning companies, the secret of Trustlook lies in the continuous learning and self-improvement characteristics of AI. “Whether it is a real threat or a false alarm, this system will always adapt to these new information and generate new response models.”

Chronicle is also a California-based company that uses machine learning to combat cybercrime. It was born as a small branch of Google in 2016, and in 2018 it has become an independent company under Alphabet. Unwilling to disclose the company’s financing or valuation, the company’s spokesperson told CNBC that this unique relationship with Alphabet allows the company to obtain more data.

Although there are many rising stars, Darktrace is still a pioneer in this field, and has really deployed its own systems in many companies. An industry insider revealed to CNBC that Darktrace has established many key technologies and capabilities, as well as a good reputation. In this market, it is the leader. Whether as an independent company or as part of a larger company, Darktrace will be better than other independent companies in the industry for a long time.

The network security market will usher in huge growth in the decade after 2020

In 2014, the number of industry vacancies worldwide reached one million, and by 2021 it will reach 3.5 million. AI is the only way to fill this gap. “We expect the future cybersecurity unemployment rate will remain at zero, and the use of AI will accelerate.” Morgan told CNBC.

The future will not only stop at detecting threats, AI will also be able to actively respond to these threats. At this point, Darktrace has already started. Two years ago, it launched a product capable of “responding to threats”. There are currently hundreds of users.

In the next step, it is not just “responding”, but AI directly and threateningly. “We are working hard on research and development, and plan to call it an automatic response system, just like we call an autonomous car.” At that stage, people can rest assured that AI technology can solve everything without any human involvement.