AI helps hackers attack the network, experts suggest solutions

Risk of data theft after just over an hour

In the era of artificial intelligence (AI), the question for many businesses is how to exploit this technology while still controlling cybersecurity risks.

AI is opening up opportunities to improve productivity, automate processes and support decision making. However, this technology can also be exploited to increase the speed, scale and sophistication of cyber attacks.

At the Ignite on Tour Vietnam event organized by Palo Alto Networks, experts said that AI is fundamentally changing the attack surface of businesses.

If previously hackers had to spend a lot of time scanning and exploiting vulnerabilities, AI can now help automate many steps, from collecting information, detecting weaknesses to creating highly convincing fraudulent content.

According to a study by Unit 42, 25% of the fastest intrusions now take just over an hour to achieve their data theft target, a sharp decrease from the 4.8 hours previously recorded. This shows that the time for businesses to detect and handle problems is getting shorter and shorter.

In Vietnam, this challenge becomes more noticeable as the process of digital transformation and AI application takes place strongly. According to data released at the event, 2025 recorded about 552,000 cyber attacks, while more than half of the organizations surveyed said they had been impacted by cyber security incidents.

Mr. Siddharth Deshpande, Director of Industry Solutions for Asia-Pacific and Japan, Palo Alto Networks, said that advanced AI systems will continue to accelerate this trend. According to him, threat actors can use AI to detect and exploit vulnerabilities at “machine speed.”

Experts say that AI also makes online fraud more difficult to identify. Many scam scenarios today not only rely on technology, but also exploit human psychological factors. Fraudsters often create emergency situations, putting pressure on victims so that they do not have enough time to verify information.

Therefore, the cybersecurity challenge in the AI ​​era lies not only in technical systems, but also in how businesses and users identify, evaluate and respond to risks.

Security must accompany the AI ​​application process

Faced with the increasingly rapid pace of attacks, many experts believe that businesses need to shift from an incident handling approach to proactive prevention. AI can also be used in monitoring activities, detecting anomalies and supporting real-time risk assessment.

One of the current challenges is that the security system in many organizations is still fragmented, using many separate tools, causing data and warnings to be fragmented. This can increase operating stress and prolong troubleshooting time.

Experts recommend that businesses build a more consistent security architecture, integrating cybersecurity right from the design and deployment of AI systems, instead of just adding layers of protection after the system has been put into operation.

Besides technological factors, user awareness is still an important link in ensuring digital safety. In the context of increasingly sophisticated forms of AI fraud, verifying information, especially in response to urgent or unusual requests, becomes more necessary than ever.

Experts say AI not only increases risks but can also become an effective defense support tool if managed properly. This technology has the ability to help detect threats early, reduce the load on operations teams and improve incident response capabilities.

As AI is increasingly integrated into production and business activities, cybersecurity is no longer a separate issue for the information technology department but becomes a part of the business's development strategy. The higher the pace of innovation, the more urgent the need for system and data protection capabilities becomes.