Protecting the truth - upholding trust in the digital society.

When the natural barrier of truth is broken

For much of human history, truth had a natural advantage: it was tied to reality, to witnesses, to physical evidence. Creating compelling information required significant effort, time, and often left traceable evidence. AI has fundamentally changed that. Today, with just a few lines of code, an individual can create dozens of versions of the same story, each meticulously tailored to different audiences. Images can be created without cameras. Videos can be produced without real events. Voices can be copied without the presence of a real person.

Illustration photo: baochinhphu.vn

Associate Professor Tran Quang Dieu, Director of the Center for Technology and Digital Transformation at the Ho Chi Minh National Academy of Politics , affirmed: “AI-generated content has broken the traditional link between information and reality. When the cost of creating fake content is almost zero, the quantity and speed of its spread become the biggest threat to society's ability to verify information.”

This observation explains why recent fake content is no longer rudimentary. It is presented coherently, structured, with sources cited, and even in the style of political journalism. This "outwardly decent" appearance lulls many people into a false sense of security. Notably, deepfake—a technology once considered a technical demonstration—has reached a new level, being used to impersonate personnel, distort statements, and misrepresent the image of leaders, especially during important political events. This is no longer just a matter of personal insults, but a matter of maintaining social stability and public perception.

Recent events have shown the emergence of manipulated videos attributing statements to leaders at various levels on sensitive issues such as personnel matters for the 14th Party Congress, the reorganization of local government at both levels, land policies, salary reform, and even foreign affairs and defense. Notably, some videos use AI-generated voices with relatively natural intonation, combined with real footage from official conferences, leading many to believe they are genuine speeches. Authorities promptly verified and confirmed these as fake products, and dealt with those behind them. However, it is noteworthy that before being exposed, these contents spread rapidly on social media, creating "undercurrents" of public opinion, especially among a segment of users who are skeptical or lack background information. This shows that even in a short period of time, fake news can damage social trust if not identified and combated promptly.

Perception manipulation: From saying false things to leading falsely.

While misinformation was previously identified by simply stating falsehoods, a far more dangerous phenomenon today is the practice of partially stating facts to arrive at false conclusions. Media experts call this "cognitive manipulation"—a process that doesn't force the recipient to believe, but rather makes them think the conclusion they arrive at is the result of independent thought.

According to Associate Professor Dr. Nguyen Quang Dong, Director of the Institute for Research on Policy and Development of Communications: "The biggest threat in the digital space today does not lie in individual pieces of fake news, but in deliberately designed content chains that manipulate public emotions and reasoning according to a predetermined scheme."

This reveals three common manipulation techniques that can be clearly identified in many recent cases: First, intentional context-adding. A truthful statement is taken out of context; a true decision is placed alongside an unrelated consequence. Viewers still perceive it as "real," but the logical flow has been distorted. Second, shifting the focus from reason to emotion. Instead of debating right and wrong, content is designed to evoke outrage, doubt, and anxiety. When emotions prevail, critical thinking ability is significantly diminished. Third, creating a sense of social consensus. Hundreds, even thousands, of accounts with similar behavior comment, share, and agree automatically, creating the impression that "this is the opinion of the majority." In social psychology, people tend to believe what they think many others also believe.

Several recent incidents involving the distortion of major Party and State policies clearly demonstrate the tactic of "saying part of the truth to mislead the whole." For example, when policies on streamlining the administrative apparatus, reorganizing administrative units, and adjusting social security policies were issued, articles and videos appeared that used accurate data and quoted documents, but deliberately juxtaposed them with negative interpretations, attributing motives, or making unfounded predictions of consequences. These contents often didn't call for direct opposition, but instead sowed doubt: "Will the policy serve the interests of a group?"; "Will the people be left behind?"; "Will the reforms erode core values?"... This seemingly "neutral" way of framing the issue is very dangerous because it doesn't deny the truth, but distorts the understanding of the truth. When disseminated continuously and systematically, these doubts accumulate, creating a sense of social insecurity.

Lieutenant General, Dr. Le Van Hai, Deputy Political Commissar of the National Defense Academy, once warned: In modern warfare, cognitive attacks are often more effective than physical attacks. When people doubt their own values, all other strengths weaken.

The problem isn't the information, but the belief.

In modern society, completely eliminating misinformation is impossible. The question is: Does that society still believe in its own self-regulation? When people begin to doubt all sources of information, even official ones, it is a sign of a deeper crisis. Trust is not lost overnight, but is gradually eroded by information gaps, delayed responses, and unconvincing explanations. According to Comrade Le Quoc Minh, Member of the Central Committee of the Party, Deputy Head of the Central Propaganda and Mass Mobilization Department, Editor-in-Chief of Nhan Dan Newspaper, and President of the Vietnam Journalists Association , speaking at the press training conference on the 14th Party Congress: “In the digital age, a single step behind in information can cause the press to lose its leading role. Information gaps are where rumors thrive.”

In reality, many incidents, if reported early, fully, and with clear analysis, would not have a significant social impact due to distorted information. Conversely, silence or vague responses are often interpreted in unfavorable ways. The requirement here is not to say a lot, but to say the right thing at the right time, to the right point, and with the necessary level of persuasiveness.

Numerous cases have shown that when official information is released early, with analysis and dialogue, distorted narratives quickly lose their foothold. Conversely, cases where responses are slow, or only brief announcements are often exploited by malicious actors to exaggerate, speculate, or even fabricate new "details" to prolong the lifespan of fake news. This reality places higher demands on mainstream media: not just reporting, but explaining policies, clarifying context, identifying the motives behind distortion, and helping the public understand the true nature of the issue.

Law enters the realm of artificial intelligence.

The National Assembly's passage of the Artificial Intelligence Law marks a significant shift: from identifying risks to establishing control mechanisms. The law approaches AI not merely as a technological achievement, but as a factor with profound social impact that needs to be regulated by law. For the first time, issues such as the responsibility of AI users, the classification of AI system risks, the obligation for transparency, and mechanisms for handling harmful behavior are systematically addressed. The core of the Artificial Intelligence Law is establishing accountability. Technology is not ownerless; when AI causes harm, there must be a legally responsible party.

This is especially important in the context of increasingly sophisticated AI-generated fake content. The law provides a basis for prosecuting not only those who disseminate it, but also the process of creating and using AI systems for harmful purposes. However, the law is only a framework; its effectiveness depends on enforcement capacity, inter-agency coordination, and the level of social consensus.

The passage of the Artificial Intelligence Law is not a spur-of-the-moment response, but rather the result of a growing awareness of the social risks posed by technology. Recent handling of AI-generated forgery cases demonstrates that without a clear legal framework, assigning responsibility is extremely difficult, especially in cases where content is created abroad but disseminated domestically. The AI Law enables authorities to be more proactive in preventing, early detecting, and addressing at its root acts of exploiting technology to harm national interests and the legitimate rights and interests of organizations and individuals.

Preventing cognitive manipulation in the AI era cannot rely on a single solution. A comprehensive strategy is needed, based on four pillars: First, proactive and transparent information; avoiding public speculation and prolonged gaps in communication. Second, mainstream media plays a leading role: not chasing sensational news, but clarifying the essence, especially in complex and sensitive issues. Third, people-centered: officials, party members, and the public must have high information discipline, not sharing based on emotions, and not inadvertently becoming links in the spread of misinformation. Fourth, law and technology go hand in hand: applying AI to detect deepfakes, trace manipulation networks, and strictly punish violations...

In a world where technology can create almost anything, the resilience of a nation is not determined by how powerful its algorithms are, but by its ability to uphold truth and social trust. Truth does not protect itself; it needs to be spoken promptly, protected by institutions, and reinforced by political fortitude.