📋 Tender Overview
On November 27, continuing the program of the 10th session, the National Assembly discussed the Artificial Intelligence (AI) Law project - a law expected to create a legal framework for AI development, application and governance in Vietnam. The National Assembly discussed the Law on Artificial Intelligence project. Expanding the testing space. Commenting on the discussion, delegate Hoang Minh Hieu (Nghe An) said that through this draft, we have clearly defined the way Vietnam approaches artificial intelligence, which is to both create friendliness for the development of AI and have regulations to minimize the risks that AI brings. The delegate emphasized that the controlled testing mechanism (sandbox) is a key solution to remove legal barriers, allowing AI to develop within a limited scope to collect data, evaluate effectiveness and promote long-term reform. According to the delegate, AI is a very new issue, involving many fields. If we only rely on the current legal framework, we cannot create conditions for AI to develop. Therefore, regulations on controlled testing mechanisms must be central to the legal framework on artificial intelligence. He proposed that the drafting agency research and amend Clause 2, Article 20 in the direction of clearly identifying organizations and individuals participating in controlled testing that are exempted or reduced from a number of obligations to comply with the provisions of this law and other relevant laws. Currently, the draft only exempts or reduces compliance obligations within the scope of the AI Law, while AI testing often involves many other laws, such as the Road Traffic Law, the Personal Data Protection Law, aviation law... Delegates cited practical examples in Japan and China: Self-driving car testing is allowed to use image data without blurred faces to train algorithms; Self-driving aircraft testing is exempt from some aviation safety rules. Therefore, the draft needs to clearly identify organizations and individuals participating in testing that are exempted or reduced from a number of obligations not only of the AI Law but also of related laws. Regarding Clause 5, Article 20, delegates proposed adjustments in the direction of the AI management agency presiding and inter-sectoral coordination in receiving and appraising testing dossiers. With specialized AI systems, such as medical image diagnosis, the role of the Ministry of Health in the monitoring and evaluation process is indispensable. Delegate Nguyen Tam Hung (HCMC) also expressed high approval for the need to promulgate the Law on Artificial Intelligence, considering this a strategic step to assert digital sovereignty, promote innovation, protect human rights and national security. Regarding the controlled testing mechanism for artificial intelligence, the delegates suggested The drafting board considered expanding the scope of testing for startups and research institutes according to a flexible testing mechanism approach, in which testing results are considered as shortened criteria when assessing conformity. This approach helps shorten the product development life cycle, encouraging Vietnamese engineers and businesses to compete with foreign technology corporations. Regarding the development of artificial intelligence human resources, delegate Nguyen Tam Hung assessed that the draft law has opened up a very correct approach. However, the delegates asked the Drafting Committee to consider emphasizing the responsibility of connecting educational institutions, businesses, and state agencies to form an ecosystem for using AI human resources. In addition to technical training, there needs to be a training program on privacy, cybersecurity and technology ethics for AI engineers and administrators to ensure AI development in a humane, sustainable and responsible manner to the community. Delegate Tran Khanh Thu (Hung Yen) raised discussion opinions Clarifying the scope of regulation and risk management mechanism in the health sector. From the perspective of the health sector, delegate Tran Khanh Thu (Hung Yen) said that AI is becoming an important tool in diagnosing, treating and managing patients, but practice also poses great risks if used incorrectly. The delegate cited the case of a 55-year-old patient in Ho Chi Minh City hospitalized in a compromised condition. Brain and stroke cases are caused by family members using AIs diagnosis results to buy medicine. The case raises legal questions about the responsibility of drug sellers and the management mechanism for AI applications in healthcare. According to delegates, current laws such as the Law on Medical Examination and Treatment or the Pharmacy Law do not have specific regulations on AI products in medicine, nor do they clearly define legal responsibility when AI makes errors. Therefore, the delegates proposed adding separate provisions to manage medical AI, especially related to risk classification, health data security and supplier responsibilities. Regarding the classification of artificial intelligence system management by risk, the delegates suggested that the drafting agency should consider having an illustrative appendix of risk levels similar to the EUs with quantitative criteria such as level of autonomy, user scale affecting basic rights or sensitive areas such as education, healthcare, justice or infrastructure, employment... Thu GiangSee moreSee more
