Hanoi (VNA) – As artificial intelligence (AI) and deepfake technology make online content increasingly sophisticated and difficult to verify, Vietnam is stepping up efforts to build a “digital shield”, strengthen the legal framework and proactively provide authoritative information to safeguard national digital sovereignty.
Efforts to build a “digital shield”
Resolution No. 57-NQ/TW, issued by the Political Bureau on December 22, 2024, on breakthroughs in science – technology development, innovation and national digital transformation, has driven significant progress in institutional reform, digital infrastructure development, public services and the digital economy. However, the digital transformation process also presents growing cybersecurity risks.
AI, big data and deepfake technology are scientific and technological achievements in themselves. The risks arise from how they are used. When exploited for fabrication and distortion, they can significantly amplify threats in cyberspace.
Le Thanh Tung, head of the Communications and Knowledge Dissemination Department under the Vietnam Union of Science and Technology Associations (VUSTA), said AI can now generate large volumes of text, images, audio and video within a very short time. Deepfake technology can produce highly convincing images and voices, creating “fake evidence that appears real” and making it difficult for users to verify information.
Big data and content-distribution algorithms can also target groups based on their psychology, interests and circumstances. False information that appeals strongly to emotions can spread rapidly and repeatedly, creating the impression of an accepted truth.
This requires earlier detection, faster verification and more timely responses, alongside proactive provision of accurate and well-grounded information. Information gaps must not be allowed to become fertile ground for fake news and distortions, Tung said.
An effective “digital shield” should have three key capabilities: early warning to identify unusual topics, sources and dissemination patterns, as well as signs of fabrication or coordinated activity; rapid verification through coordination among authorities, databases, experts and the press; and proactive communications, providing accurate, accessible and multimedia content to the right audiences once information is verified.
AI can also help analyse information needs, identify knowledge gaps and support the presentation and distribution of reliable content. However, Tung stressed that all content must be verified and subject to human responsibility rather than left entirely to algorithms.
Resolution No. 57-NQ/TW demands the mastery of technology and development of digital infrastructure, data and innovation capacity. Meanwhile, Directive No. 57-CT/TW, issued by the Party Central Committee's Secretariat on December 31, 2025, calls for stronger cybersecurity, information security and data security.
The digital shield, therefore, must form part of national digital capacity rather than being merely a standalone software solution.
Clarifying legal responsibility for AI-generated content
Vietnam has made important strides in developing its legal framework for the digital environment. The Data Law, which took effect on July 1, 2025; the Law on Personal Data Protection, effective from January 1, 2026; and the Cybersecurity Law, effective from July 1, 2026, provide an important legal foundation for managing technological risks, protecting data, defining the responsibilities of relevant parties, and handling violations in cyberspace.
However, as technologies such as deepfakes evolve rapidly, legislation must keep pace and provide clear, enforceable and sufficiently deterrent mechanisms. Priority should be given to adaptable legal frameworks that establish clear principles, responsibilities and risk thresholds while allowing for rapid technical updates, controlled testing and inter-agency coordination.
Dr. Nghiem Vu Khai, former Vice President of the VUSTA, noted that people increasingly face cross-border scams, impersonation of state agencies, fraudulent electronic contracts and financial investments, cryptocurrency-related fraud, AI-generated impersonation and personal data theft. In many cases, victims do not intentionally violate the law but lack timely access to authoritative legal advice.
He proposed prioritising the development of a national AI-powered digital legal assistance system to help organisations and individuals access official legal information, seek advice and receive timely warnings of legal risks. A national AI system should also support legal dissemination and education while ensuring up-to-date and secure legal data, personal data protection and quality control over AI-generated content.
Sharing Khai’s opinions, Tung said legal responsibility for AI-generated content should also be clarified, covering model providers, distribution platforms, deploying entities, content creators and distributors, particularly when fabricated content harms public order and safety, the dignity and reputation of individuals, or the interests of organisations and citizens.
Legal improvements must protect the rights and legitimate interests of people and businesses, safeguard access to information and preserve room for innovation, he stressed./.
VNA
