Hanoi Convention, experts propose 'digital immunity' against AI challenges

From analyzing the impacts of AI on digital trust, experts propose a "digital immunity" approach towards a safe society in the data era.

Lieutenant Colonel, PhD, criminologist Dao Trung Hieu said that the development of artificial intelligence (AI) has been profoundly changing the way people create, receive and verify information. Technologies such as deepfake, synthetic voices or artificial images make the line between real and fake blur like never before.

Lieutenant Colonel, PhD, criminology expert Dao Trung Hieu. Photo: NVCC.

According to Europol statistics (2024), about 90% of online content today has AI intervention at different levels, of which 15% shows signs of intentional forgery.

In Vietnam, the Ministry of Public Security has recorded a series of criminal cases of using AI to impersonate relatives, bank officials or state agencies to appropriate property, with some cases causing losses of up to tens of billions of dong.

What is worrying is not only the material damage, but also the erosion of social trust, the foundation of all human relations. When the truth can be recreated, trust, which is the foundation of all social relations, has become the first target of attack.

“The above phenomena are not simply a transformation of criminal tools, but a reshaping of the space of criminal behavior, with completely new characteristics: intangible, cross-border, anonymous and self-learning,” said Senior Colonel Dao Trung Hieu.

The transformation of criminal behavior structure in the AI ​​era

Lieutenant Colonel Dao Trung Hieu said that crimes in the AI ​​era are no longer limited by physical space. Criminal acts are carried out through data models, algorithms and automated generation systems, allowing individuals or organizations to fake the authenticity of all types of information.

If in traditional crimes, physical traces (evidence, crime scenes, witnesses) are the main basis, then in high-tech crimes, digital traces are the key. AI helps criminals achieve two dangerous capabilities: synthetic reality, making the recipient believe in something that does not exist, and attribution erosion, making it difficult to locate the origin of the behavior.

From a criminological perspective, it is possible to see a change in the structure of the four elements of crime:

First, the Subject is no longer a single individual but can be an automated system, or a group of people remotely controlling it;

Second, the violated object is not only property, honor or data, but also social trust - an intangible value but of fundamental significance;

Third, the means of committing crimes are programmed, self-learning, difficult to control and can be replicated indefinitely;

Fourth, the social consequences do not stop at economic damage, but spread to psychology, morality and national security.

“Thus, artificial intelligence not only supports humans in creativity and production, but also restructures deviant behavior, making the concept of “digital criminology” a new research direction of modern criminal science ,” said Mr. Hieu.

Moving from traditional investigation to investigation in the data space

Mr. Hieu said that the emergence of AI-based crimes forces investigative agencies to innovate their working models. Instead of investigating after the event, it is necessary to investigate data in parallel, collecting and analyzing the flow of information as the behavior is taking place.

Techniques such as AI trace analysis are being researched, which allow for the retrieval of specific features of each generative model similar to “digital fingerprints” in digital forensics.

In Vietnam, professional units of the Ministry of Public Security have coordinated with technology enterprises to build a database of voice, face and video recognition samples created by AI.

Along with that, the legal infrastructure also needs to be improved. It is necessary to add the crime of “using artificial intelligence to commit crimes” in the studies to amend the Penal Code, and at the same time, there must be specific regulations on the responsibilities of the owners and operators of AI systems.

What is more important is to shift from a “hunting for the culprit” mindset to a “proactive prevention” mindset. In an era where a fake video can spread faster than an official announcement, the ability to quickly respond and coordinate among all parties, including the State, businesses and the community, becomes a key factor.

Digital immunity, “antibodies” in the AI ​​era

A concept emphasized by Lieutenant Colonel, Dr. Dao Trung Hieu is “digital immunity”, which is the ability of society to self-identify and react to fake information. To build digital immunity, according to him, it is necessary to converge three groups of solutions.

The signing ceremony of the United Nations Convention against Cybercrime took place in Hanoi on October 25. Photo: Government.

First, build a flexible and evolutionary legal corridor, ensuring a balance between control and innovation. Laws should not “frame” AI, but need to be flexible to adapt to rapid changes in technology, while ensuring the principle of dual responsibility - individuals or organizations using AI must be responsible for the products they create.

Second, establish a national identification mechanism for AI-generated content, through digital origin authentication standards (AI watermark, provenance code). Digital platforms should be required to label or verify AI-generated content, helping users distinguish between real and fake right from the reception stage.

Third, education and social media are a long-term line of defense. Laws can regulate behavior, but only education can regulate awareness. Integrating knowledge about data security, fake news recognition skills, and technology ethics into general education and university curricula will create a class of citizens with “digital antibodies,” a sustainable foundation for a safe society.

According to Mr. Hieu, artificial intelligence is not only a technological advance, but also a moral and legal test of humanity. When machines can recreate the truth, people are forced to learn how to protect the only belief that cannot be programmed.

Vietnam needs to proactively take the lead by building a flexible legal framework, enhancing public-private cooperation in AI crime prevention, and promoting digital immunity education for the people.

“A safe society in the age of artificial intelligence is not the one with the strongest firewalls, but the one with the most alert citizens. If data is the fuel of the digital age, then trust is the engine. Without trust, any system collapses, no matter how advanced the technology,” the criminologist emphasized.

On October 25, the signing ceremony of the United Nations Convention against Cybercrime with the theme "Combating Cybercrime - Sharing Responsibility - Looking to the Future" took place in Hanoi. The Hanoi Convention is considered an important step in building the first global legal framework to coordinate joint efforts to respond to cybercrime. This is a historic milestone not only for the international community but also affirms the role and position of Vietnam in the international arena.

Source: https://khoahocdoisong.vn/cong-uoc-ha-noi-chuyen-gia-de-xuat-mien-dich-so-truoc-thach-thuc-ai-post2149063689.html