Combining humans and AI creates a defensive wall against cyber attacks

Reporter of Tin Tuc and Dan Toc newspaper had an interview with Mr. Vu Ngoc Son, Head of Technology Department, National Cyber Security Association, Technology Director of NCS Company about the above topic.

How is the application of artificial intelligence in the field of cybersecurity being deployed in Vietnam, sir?

AI helps detect anomalies based on log analysis, malware analysis and reverse engineering, and security vulnerability hunting.

Photo caption — Mr. Vu Ngoc Son, Head of Technology Department, National Cyber Security Association.

If only using humans, finding a cybersecurity vulnerability can take 2 weeks, but now with artificial intelligence (AI) it only takes about 5 minutes. Finding vulnerabilities is both a challenge for cybersecurity experts and a testament to their ability.

Nowadays, large enterprises have very high requirements for network security. The system testing to operation only takes about 1 week. Therefore, applying AI in the field of network security is an inevitable trend. In particular, AI is very good at automating processes and identifying user behavior.

According to you, what are the challenges when applying AI?

We are looking at AI like “blind men touching an elephant”, because each person's perspective and level of understanding on this issue is different.

Many businesses and organizations share their use of AI. Experts also have different opinions about AI, including whether it is good or bad. It depends on the approach and the answers that AI gives.

There is still no comprehensive picture of the positive and negative impacts of AI. The biggest challenge is the human factor. AI depends a lot on the ability of the user. Some experts say that AI is a “digital twin” of the user or a knowledge duplicator.

One problem today is that cybersecurity experts are not necessarily data experts, and vice versa. There are not many people in Vietnam who are good at both of these skills. I have invited many international AI experts to Vietnam to support. They all require cybersecurity experts to have a minimum knowledge of AI to be able to deploy solutions. This shows that humans still play a key role in AI applications.

In terms of technology, most of the models we are using are reused platforms from abroad. The training data is a “black box” that we do not fully understand. If we want to use it, we must retrain the AI to verify whether the results are correct or not.

Financially, we lack the resources to invest in research laboratories. In addition, it is difficult to hire high-quality experts, and there is not enough potential to maintain long-term projects. Some businesses find it interesting when testing, but when implementing in practice, they get stuck, so it is difficult to invest to the end.

Who can master artificial intelligence in this field of cybersecurity, sir?

There are two groups of players in this field. Defenders use AI to detect anomalies, develop security solutions, and automate processes. On the other hand, attackers use AI to detect and exploit vulnerabilities; test against systems; automate attack stages, and cover their tracks.

AI is important from both sides. Whoever implements AI better will dominate the game.

In fact, the attack groups are often backed by financially strong organizations, so they have the advantage. On the contrary, the defense side is also invested in AI by businesses and organizations, deploying technological solutions to improve efficiency.

Currently, we are seriously lacking in human resources in this field. The Association's 2024 survey of 3,000 businesses shows that: 20% of businesses do not have specialized cybersecurity personnel; 35% have less than 5 people.

Meanwhile, network security monitoring must be performed 24/7. With such a human resource ratio, it is very difficult to ensure the network security of that enterprise's system.

To solve this problem, what solutions do you think businesses and organizations should have?

In my opinion, using AI in cybersecurity in Vietnam requires combining humans and technology to create an effective defense wall. AI should only be fully authorized when the technology is truly excellent and the process is clear.

Therefore, first of all, it must start with humans. If humans do not master the expertise when using AI, it will inadvertently multiply the vulnerabilities.

Investing in AI must go hand in hand with investing in teams and processes. AI only has value when it is in the hands of people who know how to master the technology.

AI is not a “magic bullet” without standard processes and competent people. However, AI is becoming a competitive advantage, especially for large enterprises.

Thank you very much!

Source: https://baotintuc.vn/xa-hoi/ket-hop-con-nguoi-va-ai-tao-buc-tuong-phong-thu-truoc-cac-cuoc-tan-cong-mang-20250922233448451.htm