Friday, December 2, 2022
HomeTechnologyArtificial intelligence enters the fight against cybercrime

Artificial intelligence enters the fight against cybercrime

With systems that learn from their own experience, they manage to anticipate future attacks and save resources.

The artificial intelligence (AI) It has advanced so much that it is already part of everyday life. Thanks to thousands of traffic data collected and analyzed every day, Waze or Google Maps predict how long it will take to move from one place to another; and everything optimized with current information. Even the City Government, with its virtual robot botihe can anticipate some questions from the porteños who interact with him.

Now, AI is entering the world of cybersecurity, with the idea of ​​staying ahead of attacks while reducing resources.

“For example, we started to provide an AI-based cybersecurity service to a large company. And in just a few months, from data collection and learning, our system began to separate the real attacks from the ones that looked like attacks and were not. This is how we gain efficiency. We started out working with 16 people, basically on detection and solutions. Today we do this work with just three people”, he tells Clarion Gabriel Catropa, Technical Manager of IBM Security at a meeting on security and Artificial Intelligence at the MIT-IBM Watson laboratory of this company and the prestigious Massachusetts Institute of Technology (MIT) in Cambridge, Massachusetts, United States.

“The three technicians who continue to work for that company are focused on solving the attacks. And the detection is now practically in the hands of artificial intelligence. With AI, the response is proactive and solutions are accelerated”, explains Catropa.

The IBM X-Force Command Center bunker, in Cambridge, United States.

Cyber ​​protection is now more sophisticated for a simple reason: virtual assaults are becoming more and more difficult to detect. The ransomware (they steal files and ask for a ransom to recover them) was the main type of attack in Latin America in 2021 (29%). And phishing was the most common infection vector used in the region, accounting for 47% of attacks. The data corresponds to the X-Force Threat Intelligence Index 2022 study, by IBM. In addition, according to the same work, in Latin America the average time to identify and contain a data breach was reduced by almost 25 days (331.5 days) but is still above the global average of 277 days.

It is estimated that 61% of those who suffer a ransomware attack pay to try to recover the information. “But paying is not profitable. First of all, no one guarantees that you will recover the data, and secondly, you become a ‘paying victim’, that is, the criminals know that you pay and that is why they can try to sabotage your networks again”, indicates Catropa.

in cambridge, Clarion was able to access the IBM X-Force Command Center, one of the company’s bunkers to combat cybercrime. In the place they also do demonstrations for companies on how they deal with hacking attempts.

The amazing State Center building, in the prestigious Massachusetts Institute of Technology (MIT).

The amazing State Center building, in the prestigious Massachusetts Institute of Technology (MIT).

“For many years, the financial industry was the most attacked in the world, but in 2021 manufacturing became the most attacked industry. In addition, 60% of companies increased the prices of their products or services after a cyberattack, which contributed to the general increase in goods worldwide, “said Mike Barcomb, executive director of the IBM Security Command Center, in his presentation.

All in all, it seems that these are times of artificial intelligence, a technology that will be enhanced with the arrival of the speed of 5G. In Argentina, only 19% of companies have already incorporated AI, but 51% are beginning to implement it.


A researcher takes a break from his activity, in the MIT-IBM laboratories, in Cambridge, USA.

A researcher takes a break from his activity, in the MIT-IBM laboratories, in Cambridge, USA.

Artificial intelligence is expanding and is also being used by cybercriminals who are at the forefront of attacks. AI works by learning and, like any human, its “mind” can also be “corrupted” with unwanted information. An example of this was the Tay Chatbot, a chatbot created by Microsoft for Twitter in 2016. Tay caused controversy because she was “poisoned” and started sending offensive messages.

So-called Evades, real-world attacks on computer vision, usually for facial recognition biometrics and autonomous vehicles, were also recorded. Yes, even the AI ​​can be fooled.

In addition, AI can be used to generate automation messages for phishing.

Finally, neural networks are becoming more “refined” to be able to crack passwords. According to Catropa, a 12-character password can already be cracked in a minute.

Massachusetts, United States. Special delivery.


Look also

Recent posts