Carefully calculate when bringing AI into public management

The strong development of artificial intelligence (AI) has brought many breakthrough opportunities in state management, helping to improve management processes and reduce administrative procedures, especially in the field of public finance. However, to effectively apply AI to the current infrastructure, data management capacity, human resources and mechanisms and policies of the public sector is a problem that needs to be carefully calculated.

Many challenges

In Vietnam, the deployment of AI in state agencies has made significant progress. Notably, the Government has issued a National Strategy on AI Research, Development and Application to 2030. Many ministries, sectors and localities have initially applied AI in public management and services. These include virtual assistants (chatbots) supporting administrative procedures, facial recognition software serving security and order, and smart traffic camera systems, etc.

However, the process of applying AI in state agencies still reveals some difficulties and limitations such as limited application, problems not close to practical needs, lack of legal basis, guidance and lack of attention to AI risks. In addition, there are difficulties in financial resources, financial mechanisms, data and infrastructure, human resources... Mr. Do Quang Tuan ( Vietinbank , Hanoi branch) said that implementing AI is not easy due to data security issues. and privacy is always a top concern. The investment budget to upgrade to AI infrastructure is the biggest difficulty that banks face.

In competition management and consumer protection, Mr. Trinh Anh Tuan, Chairman of the National Competition Commission ( Ministry of Industry and Trade ), commented that competition management and consumer protection agencies are facing increasingly greater challenges. Detecting sophisticated anti-competitive behaviors, handling large numbers of consumer complaints effectively and predicting potential market trends require new tools and approaches. This is where AI can help build predictive models and risk warnings; warn consumers, detect and prevent online fraud, etc.

Mr. Tuan also pointed out that it is necessary to clearly recognize the challenges in the process of implementing AI applications in a number of fields, especially in the field of consumer protection. These are issues of data quality and availability, the complexity of AI algorithms requiring a team of highly qualified experts; ethical issues, concerns about the privacy of user data, etc.

According to Dr. Tran Van Khai, Vice Chairman of the National Assembly's Committee on Science, Technology and Environment, implementing AI in the public sector still faces many difficulties. A recent report shows that 87% of AI projects in state agencies have not been implemented in practice, and 70% of units assess the effectiveness of AI integration as low. The reasons are the lack of data and synchronous digital infrastructure, lack of high-quality AI human resources, along with an incomplete legal framework and technical standards, and financial constraints. "For example, the whole country currently has only about 300 AI experts, too few compared to the human resource needs for digital transformation," Dr. Khai emphasized.

Prioritize appropriate and scientific AI applications

According to experts, with the right strategy, reasonable investment and ensuring ethics, AI will become a driving force for national development and people's lives. To fully exploit the potential of AI, it is necessary to have a clear strategy, focusing on investing in data infrastructure and AI technology; building a team of competent experts, building a suitable legal and ethical framework. Dr. Phan Ke Son, Secretary of the Youth Union of the Vietnam Academy of Science and Technology, said that AI needs to be deployed in problems with large enough data, clear rules and bringing specific values. The digitalization process needs to be accompanied by standardization of data formats to ensure consistency and ease of integration; there needs to be a clear and synchronous process in collecting, storing and updating data. Agencies and units need to note that when using AI, data sharing must be accompanied by a security mechanism, clear access authorization; it is necessary to develop policies on privacy, data encryption and user authentication. Each agency needs to clearly identify priority problems to select and deploy appropriate AI, with a roadmap, to avoid spreading.

DO LIEN and NAM PHONG

Source: https://nhandan.vn/tinh-toan-ky-khi-dua-ai-vao-quan-ly-cong-post883787.html