Speaking at the CIO Summit 2026 held last Thursday in Ho Chi Minh City, Minh said AI may not create an entirely new attack surface, but it could significantly increase the scale, speed, and number of attacks that hackers can launch against businesses and organizations.

Security will be the hottest field over the next three years, he said, outlining VNG's view on the impact of AI on cybersecurity.

With current AI models, hackers can automate multiple stages of an attack and dramatically increase the number of attempts against a target, he said. This means offensive capabilities can expand rapidly without a corresponding increase in the number of people carrying out attacks.

Defensive capabilities face greater constraints, he added. Businesses cannot give AI the same level of autonomy when protecting critical systems, particularly in banks and organizations holding large amounts of sensitive data. Security tools must first reach a very high level of reliability before they can be allowed to take action automatically, the executive noted.

“This is an important paradox of AI,” Minh stressed. “AI is making offensive capabilities much stronger than defensive capabilities.” The risk extends beyond banks to virtually any organization.

VNG takes a bottom-up approach to AI adoption

Minh's warning comes as companies accelerate the use of AI in their operations. VNG has taken a different approach from highly regulated organizations such as banks, he said. Banks typically adopt technology from the top down, selecting solutions that have already been tested and meet strict requirements for security, scalability and reliability.

VNG instead aims to give employees more freedom to experiment with AI in their daily work. Management's role is to build the underlying foundations, including data, security and technology tools, and then encourage employees to find their own ways to use AI, he said. This includes developing internal AI agents and redesigning day-to-day workflows.

The approach reflects VNG's view that the value of AI extends beyond adding another technology layer to existing systems. Its larger potential lies in expanding employees' ability to innovate and changing how individuals work.

AI investment is also an investment in people

Minh said companies should also view AI spending differently from traditional technology investment.

In the past, large IT investments typically involved core banking systems, data centers and servers. With AI, a significant part of current spending is linked to tokens, which remains manageable compared with overall technology budgets, he said. “Investing in AI today means investing in technology and people at the same time,” Minh affirmed.

The immediate goal should not necessarily be to prove how much additional revenue AI can generate or how much cost it can cut. Companies should first build the ability to use AI across the organization, he suggested.

If employees can improve their productivity through AI, the cost of deploying the technology may remain relatively small compared with overall IT spending. Less sensitive data can be processed using commercial AI models, while critical or highly sensitive information should be handled on dedicated infrastructure.

Minh rejected the view that AI will inevitably eliminate large numbers of technology jobs. He expects the opposite could happen as AI increases productivity and allows businesses to operate at greater scale.

Demand for people with a strong understanding of technology could rise, he said, because those workers are also better positioned to use AI effectively. As AI becomes more deeply embedded in business operations, the ability to understand, deploy and control the technology could become a more important source of competitive advantage for both companies and their employees.