Vietnam develops national protection law

The concern was shared by Professor Toby Walsh, University of New South Wales (Australia) – one of the globally influential voices on AI on the sidelines of VinFuture Science Week 2025 on the afternoon of December 2.

He believes that many developing countries were physically colonized when data became the new resource. When these countries are not proactively controlled, they face ‘digital domination’.

Analyzing further, he commented that Vietnam’s values ​​and culture are different from those of Australia, China and the US. “We cannot expect technology companies from China or the US to automatically protect Vietnamese culture and language. Those things must be proactively protected by Vietnam itself,” he said, saying that building an AI law is an important step to maintain autonomy.

According to him, each country has its own values ​​and culture and needs laws to protect these values. Therefore, he was “very happy” to see Vietnam being one of the pioneering countries in building specialized laws on artificial intelligence.

Responsibilities of AI developers

According to Professor Walsh, we cannot rely on the goodwill of the market, because too large profits from AI can push businesses beyond ethical limits. The only way to ensure they always behave correctly is to apply strict regulations, so that public interests are balanced with commercial interests. “Responsible use of AI should be mandatory,” he said.

He cited the example of an AI system used to assess the likelihood of recidivism and recommend prison sentences in the US. Because it was trained from historical data, this model unintentionally reflected past racial prejudice, leading to recommendations for harsher sentences for black people. “It cannot be left to such a system to decide to imprison someone,” he said.

Unlike humans, AI cannot be held responsible when it makes mistakes. But now, when AI acts as a new “agent”, able to make decisions and take actions, if allowed by humans, the question is who will be responsible if AI makes mistakes?

Professor Walsh believes that this responsibility belongs to the companies that deploy and operate AI systems. “We must force them to bear responsibility for the consequences that these “machines” cause,” he expressed his opinion.

According to him, mental health risks from AI are not theoretical. He referred to lawsuits in the US where parents accused technology companies of causing their children to commit suicide after interacting with chatbots. The number of ChatGPT users experiencing mental health problems, even if “only a few percent”, is equivalent to hundreds of thousands of people.

For young children, he believes that lessons from social networks have not yet been fully learned. Some countries, including Australia, have age restrictions on social media use. This is something that countries that are drafting laws should consider applying to AI.

In the context that Vietnam is perfecting the AI ​​Law, he believes that many current regulations such as privacy laws and competition laws can still apply to the digital space. To avoid the downside of AI, the professor’s advice is to invest in people, improve skills and understanding of AI for all people, in parallel with developing the domestic ecosystem. “Be proactive. Don’t wait for other countries to transfer technology or direction,” he suggested.

Vietnam’s draft AI Law currently outlines eight basic principles, the most important of which is “human-centered”, ensuring AI respects human dignity, privacy and cultural values. The draft law also classifies this technology according to four risk levels. Unacceptable risk groups, including systems that manipulate behavior, attack vulnerable groups, produce fake content that endangers national security or support illegal acts, are proposed to be banned from development and circulation.

According to the Ministry of Science and Technology, the spirit of the law is not to tighten management, but to create an “open and safe” legal corridor to promote development. Along with perfecting the law, Vietnam is also promoting investment in computing infrastructure, financial resources and AI human resource training.