In that context, artificial intelligence (AI) has emerged as one of the powerful and effective tools that promises to revolutionize tax administration and help tax authorities effectively deal with these challenges.
Big data analysis, automating tax management steps
Since 2023, the tax industry has begun applying Big Data and AI in invoice management, quickly detecting, warning and promptly preventing acts of invoice fraud, tax refund fraud, tax evasion, creating a healthy and equal business environment for the business community; at the same time, building and putting into operation the application of digital maps of business households, digital maps of mines, resources and minerals.
One of the most important applications of AI in the tax sector is automated data analysis. Machine learning algorithms can be trained to analyze large volumes of transaction data, automatically classify taxpayers, determine tax liabilities, classify goods and services, and even detect potential tax evasion. This not only significantly reduces administrative time and costs, but also improves the accuracy and efficiency of the tax collection process.
With a huge amount of data collected from electronic invoices, electronic tax declaration and payment information (eTax), the tax industry is applying artificial intelligence (AI) technology and big data analysis (Big Data) to improve management efficiency, detect and prevent tax evasion and tax fraud.
From building a large, centralized database connected to the national population database (according to Project 06/CP), the tax industry has been applying Big Data and AI to manage invoices, quickly detect and prevent invoice fraud, tax refund fraud, tax evasion and avoidance, creating a healthy and equal business environment for the business community.
For taxpayers, AI also brings many benefits.
Sharing with Lao Dong Newspaper, Mr. Vu Manh Cuong - Deputy Director General of the General Department of Taxation, concurrently Director of the Hanoi Tax Department - confirmed that the Hanoi Tax Department is currently the unit assigned by the General Department of Taxation to pilot the application of the Virtual Assistant model to support taxpayers.
“Since the beginning of 2024, the implementation team at the Hanoi Tax Department has collected data sources of legal documents on tax policies and related legal policies (with 100 specialized laws), tax administrative procedures, the system of documents responding to taxpayers... and has been refined and combined with artificial intelligence technology. Currently, the application has been put into effective use, answering over 30,000 questions from taxpayers, automatically, quickly, accurately with detailed, easy-to-understand content, integrating current administrative procedure forms, intuitive and easy-to-understand instructional clips".
Reduce fraud
Risk analysis and fraud detection is also an area where AI has a significant contribution. Machine learning algorithms can analyze complex data patterns to identify suspicious transactions, signs of transfer pricing, tax evasion, and other fraudulent behavior.
For example, AI can detect cases of under-declaring goods to evade import duties or cases of using fake addresses to avoid paying taxes. Using AI in risk analysis helps tax authorities focus resources on high-risk cases, improving the effectiveness of inspection, examination and control against tax fraud.
To date, the e-invoice system has received and processed over 11 billion invoices, starting from 2022, when tax authorities apply e-invoices nationwide. With this huge amount of data, only AI can “read” it effectively and provide useful information to tax authorities.
With a large database of electronic invoices, the General Department of Taxation also researches the application of a number of new technologies to analyze data such as natural language processing solutions to identify names of goods and services on electronic invoices; find unusual prices of goods; detect chains of goods and services purchased and sold through many intermediaries to identify cases where taxpayers show signs of risk in tax law compliance.
However, Mr. Do Hoang Anh Tuan - Deputy Director of the Digital Technology Center, Department of Digital Transformation - Ministry of Information and Communications - also stated that the application of AI in e-commerce tax collection also poses some challenges.
“First, the technology infrastructure is still limited and does not meet the requirements well. Although the Big Data system has been deployed, the technology infrastructure is not strong enough to handle the large and complex data volume from digital economic activities, leading to overload or delay in data processing. Cybersecurity risks increase as the number of transactions and data is digitized, while security measures are not really comprehensive.
Second is the challenge of technology human resources due to the lack of specialized personnel. The tax industry needs many experts with deep expertise in technology such as AI, Big Data, Blockchain, etc. to effectively implement digital transformation solutions. However, the competition for human resources between the public and private sectors makes it difficult for the tax industry to attract leading technology experts.
Experts also point out other issues such as data quality. Machine learning algorithms are only effective when trained on high-quality data. A lack of complete, accurate, and consistent data can affect the reliability of the analysis results.
Not to mention that machine learning algorithms often operate as “black boxes”, making it difficult to understand how they make decisions. This can make it difficult to audit and monitor the system’s operations, as well as cause suspicion and distrust from taxpayers.
To overcome these challenges, close cooperation between tax authorities, technology experts and other stakeholders is needed.
Investing in technology infrastructure, developing common data standards, and building clear regulatory frameworks are important steps to ensure the effective and sustainable application of AI in e-commerce tax collection. Training human resources is also an important factor, helping tax officials grasp new technologies and apply them effectively in their work.
Mr. Mai Xuan Thanh - General Director of the General Department of Taxation - Ministry of Finance - informed: "AI can automatically classify goods and services based on product and service descriptions, helping to accurately determine the tax rate applicable to each type of goods and services. Furthermore, AI can be used to analyze and determine the sales transactions and revenue of individuals and e-commerce businesses, helping tax authorities identify areas and tax losses and plan to collect and prevent budget losses more effectively.
Since the tax authorities implemented electronic tax (eTax) and electronic invoices, not only have tax payment procedures for taxpayers been simplified, quick and convenient, but a transparent tax management system has also been created with the help of AI. The combination of electronic tax and electronic invoices has created an effective tax management system, helping taxpayers best fulfill their tax obligations, making an important contribution to collecting the correct and full amount of tax.
