Instructions on using AI chatbots to assist officials and civil servants in their work.

Illustrative image. (Source: Ministry of Science and Technology )

To fulfill its function of state management over scientific research, technological development, and innovation, and to meet the practical requirements of applying Artificial Intelligence (AI) technology in state agencies, the Ministry of Science and Technology has issued guidelines on some general principles for civil servants, public employees, and workers in state agencies using large language models (AI chatbots) for work purposes.

This guide aims to provide general, easy-to-understand principles for civil servants, public employees, and workers in government agencies to use large language models (AI chatbots) to support their work effectively and safely.

Organizations and units that need to apply AI chatbots in their work are encouraged to proactively learn about and use this document to implement and disseminate it within their organizations in a way that suits their specific needs. This document helps ensure that the application of AI is safe and delivers practical benefits.

Director General Mai Xuan Thanh, along with Vice Chairman of the Hanoi People's Committee Ha Minh Hai, Deputy Director General Vu Manh Cuong, President of the Vietnam Tax Consulting Association Nguyen Thi Cuc, and Deputy Director of the Hanoi Tax Department Phan Tien Hoa, performed the ceremony of pressing the button to launch the "Virtual Assistant to Support Taxpayers" application.

The Ministry of Science and Technology clearly states that the principles for providing data and using AI chatbots are: No data or documents classified as state secrets, as well as sensitive information such as personal information, internal documents, or data that could affect security or the organization, should be shared with the AI chatbot. Ensure that the data or information provided to the AI chatbot does not contradict Vietnam's historical traditions, culture, ethics, geography, political views, territorial sovereignty, or customs. Furthermore, do not use content that infringes on copyright, intellectual property rights, or other related rights.

When using AI chatbots, avoid creating accounts with usernames or passwords that are the same as official accounts (work email accounts, internal systems, etc.) to ensure information security. Use AI chatbots cautiously, effectively, and responsibly. Also, carefully check and evaluate the results provided by the AI chatbot before use. Ensure the information is accurate, relevant to the work context, and does not cause misunderstandings or errors when applied. Do not rely entirely on AI chatbots; combine their use with personal knowledge and experience to make informed decisions.

For agencies and units, it is necessary to disseminate information and guide officials, civil servants, and employees on the use of AI chatbots in accordance with the functions, duties, powers, and specific characteristics of the agency or unit. Emphasis should be placed on ensuring cybersecurity.

The responsibility for using AI chatbots rests with the head of the agency or organization. They are responsible for managing and securing the AI chatbot account; in the event that the agency or unit's account is compromised or impersonated, the agency or unit must notify the AI chatbot deployment unit for timely action.

Access management and permissions in AI chatbots must be consistent with the functions, tasks, and authority of the participants using the AI chatbot.

When officials, civil servants, or public employees resign or transfer to another position, the agency or unit must retrieve the security key (if any) or delete the AI chatbot user account to ensure security and strict management.

Gradually build your own knowledge database, and then establish specialized chatbots to serve the operations of your agency or unit.

Regularly monitor and promptly assess the usage of AI chatbots by officials, civil servants, and employees under their management. The unit deploying the AI chatbot is responsible for providing detailed instruction manuals and regular technical support to ensure smooth and safe usage.

For officials, civil servants, and employees, it is necessary to self-manage and take responsibility for protecting the personal information declared on the information platforms and systems of the agency or unit; they must not disclose login accounts or make unauthorized connections between the agency's information platforms and systems and the AI chatbot.

When using chatbots in general, and AI chatbots in particular, at public internet access points, absolutely do not enable password storage during use. Do not provide personal data, identity information, or other sensitive information of yourself or others to the AI chatbot. Do not use chatbots in general, and AI chatbots in particular, for illegal purposes, violating public morals, or engaging in activities such as hacking, creating deepfakes, or fraud.

During the implementation process, if any difficulties, obstacles, or major problems arise while using AI chatbots, agencies and units should contact the National Digital Transformation Department, Ministry of Science and Technology, for support.

Reportedly, the Ministry of Science and Technology is developing a national AI transformation program to apply AI in all fields and build a shared AI computing infrastructure.

The national AI transformation program must reflect a "nationwide" and "comprehensive" approach, aiming for every citizen and public official to have a virtual assistant.

Accordingly, the deadline for submitting the National AI Transformation Program to the Prime Minister for promulgation is May 2025, and a virtual assistant must be available for citizens by November 2025.