Discussion on the Application of Internet of Things and Artificial

Intelligence in Tax Administration

Runsheng Gao

Business College, East China University of Science and Technology, 200237, Shanghai, China

Keywords: Internet of Things, Artificial Intelligence, Tax Management, Digital Transformation, Risk Prediction.

Abstract: This study explores the application and synergistic effects of the Internet of Things (IoT) and Artificial

Intelligence (AI) in tax management. The research addresses the challenges faced by traditional tax

management systems, such as data lag, high costs of manual auditing, and limited accuracy, which are

exacerbated by human intervention. The integration of IoT and AI is proposed as a core driver for the digital

transformation of tax management. IoT enables real-time data collection through smart sensors, enhancing

data accuracy and monitoring capabilities, while AI leverages machine learning and natural language

processing to analyze tax data, identify patterns and anomalies, and support intelligent risk assessment and

decision-making. The study employs a combination of theoretical analysis and case studies to examine how

IoT and AI can optimize tax data collection, improve tax risk prediction, and address the challenges faced by

enterprises in adopting these technologies. The findings suggest that the integration of IoT and AI significantly

enhances the efficiency, accuracy, and transparency of tax management, reducing compliance risks and

operational costs. However, challenges such as high implementation costs, data privacy concerns, and

technical adaptability remain. The results of this study provide new insights for the digital transformation of

tax management.

1 INTRODUCTION

Against the backdrop of an increasingly complex

global economy, traditional tax administration is

beset by problems such as lagging data, high costs

and low efficiency of manual review, and limited

accuracy. Manual intervention further escalates the

risk of errors, making improvement an urgent need

(Song, 2023). The Internet of Things (IoT) and

artificial intelligence (AI) have emerged as the core

drivers for the digital transformation of tax

administration. IoT enables real-time data collection

through smart sensors, enhancing data accuracy and

the ability to monitor corporate behavior, and

improving decision-making efficiency. AI, on the

other hand, leverages machine learning and natural

language processing to deeply analyze tax data,

identify patterns and anomalies, support intelligent

risk assessment and decision-making, reduce human

errors, and significantly enhance efficiency (Wu,

2024). However, challenges such as information silos

and poor technological adaptability still impede the

https://orcid.org/0009-0005-9858-8395

transformation process. Therefore, effectively

integrating IoT and AI to collaboratively optimize tax

administration is the key to achieving intelligent and

digital transformation.

This research mainly focuses on the application

and synergy of IoT and AI in enterprise tax

management. With the above research background,

this paper raises the following three specific questions:

First, how can IoT optimize tax data collection and

enhance the real-time and accuracy of tax

management? Second, how can the application of AI

in tax data analysis and decision-making improve tax

risk prediction and management? Third, what are the

main challenges that enterprises face when adopting

these technologies?

This research will explore the possibility of IoT

technology, AI technology and tax management from

both theoretical and data perspectives, and will also

discuss specific cases. The significance of this

research lies in promoting the intelligence of tax

management. Through IoT and AI technologies, it

aims to enhance the real-time and accuracy of tax data

Gao, R.

Discussion on the Application of Internet of Things and Artiﬁcial Intelligence in Tax Administration.

DOI: 10.5220/0013860200004719

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on E-commerce and Modern Logistics (ICEML 2025), pages 751-756

ISBN: 978-989-758-775-7

751

and reduce tax compliance risks. This research will

combine the latest application cases of AI and IoT to

provide new perspectives and solutions for

enterprises and academia, and contribute to the

modernization of tax management.

2 THEORETICAL BASIS

2.1 IoT and Tax Administration

This section focuses on the crucial role of IoT in tax

administration, and it is mainly divided into three

aspects. First, real-time data collection. IoT achieves

monitoring of the entire logistics process through

GPS and wireless sensors, providing real-time data

support to ensure accurate collection of value-added

tax, consumption tax, etc. (Yan, 2021). Second,

intelligent invoice management. By integrating IoT

with blockchain technology, an intelligent invoice

system is created. Through RFID tags, the entire

invoice process can be monitored, ensuring

authenticity and compliance, and enhancing

management transparency and review efficiency (Shi

et al, 2024). Thirdly, tax management for fixed assets.

The IoT utilizes RFID technology to achieve precise

tracking and status monitoring of enterprise fixed

assets, ensuring accurate depreciation calculation,

optimizing asset management, and reducing tax

compliance risks (Chen,2024).

2.2 AI and Tax Administration

The application of AI in enterprise taxation mainly

includes the following types. The first is tax risk

prediction, where machine learning identifies

potential tax compliance risks for enterprises; the

second is intelligent tax declaration, which

automatically fills in and optimizes tax strategies; the

third is the identification of false transactions, where

AI combined with IoT data detects abnormal

transaction patterns. At the same time, the

development and introduction of AI robot customer

service facilitate more convenient communication

between tax authorities and enterprises, saving human

and material resources (Zheng et al., 2022).

It employs three core technologies in tax

administration. The first is natural language

processing (NLP), which is used to automatically

identify tax compliance issues; the second is machine

learning, which is used for tax data analysis and

prediction of tax risks; the third is computer vision,

which is used for automatic invoice review and

identification of forged documents (Merola, 2022).

2.3 Analysis of the Possibility of

Combining IoT Technology, AI

Technology and Tax

Administration

2.3.1 Theoretical Support

Technology Acceptance Model (TAM model) is a

theoretical model used to explain how users accept

and utilize new technologies. It mainly focuses on the

influence of perceived usefulness and perceived ease

of use on users' behavioral intentions. According to

the TAM model, the acceptance of new technologies

by users is determined by the perceived usefulness and

perceived ease of use of the technology (Almahri et al.,

2025). In tax administration, IoT and AI have

enhanced tax efficiency and accuracy by providing

real-time data and intelligent analysis, meeting the

demands of tax officials and enterprises for

technology.

Structure-Conduct-Performance (SCP framework)

is an industrial organization analysis model that is

used to study the interrelationships among market

structure, enterprise behavior and market

performance (Panhans, 2023). According to the SCP

framework, the structure, behavior and performance

of the market determine the competitive situation of

the market. The combination of the IoT and AI

technology optimizes the tax management structure,

enabling the tax department to conduct tax

monitoring and risk management more precisely and

enhancing the performance of tax management.

2.3.2 Data Analysis

The combination of IoT and AI in tax management

demonstrates remarkable advantages. Through real-

time and precise data collection by IoT and deep

analysis and intelligent decision-making by AI, it not

only enhances the timeliness, efficiency and

transparency of tax management, but also reduces data

errors and tax risks, promoting the transformation of

the tax system towards intelligence and automation.

This technological integration provides important

opportunities for improving tax compliance,

enhancing public trust, and promoting cross-

departmental data cooperation. However, its

application also faces high technical implementation

costs, insufficient technical adaptability, and

disadvantages such as data privacy and security.

Moreover, immature technology, system

compatibility issues, legal and policy restrictions, and

potential stability threats due to over-reliance on

ICEML 2025 - International Conference on E-commerce and Modern Logistics

752

technology may all limit its comprehensive promotion

and application effectiveness.

3 APPLICATION SCENARIOS

AND CASE ANALYSIS OF IOT

AND AI IN TAX

ADMINISTRATION

3.1 Specific Applications and Analysis

of the IoT

3.1.1 The Brasil ID System

The Brasil ID system ("Brasil ID") is an innovative

project jointly initiated by the Ministry of Finance of

Brazil, the National Tax Bureau, and the state finance

departments in 2007. Its aim is to optimize tax

management in interstate trade through Radio

Frequency Identification (RFID) technology. This

system addresses issues such as low regulatory

efficiency and severe tax fraud in the ICMS

(Interstate Sales Tax) for interstate goods and services

circulation in Brazil. By using RFID tags to store

information about goods and automatically collecting

data through state interstate highways antennas and

comparing it with tax records in real time, it enables

compliance verification. As of 2023, the system has

been implemented in 13 states, significantly

enhancing transportation efficiency (reducing

approximately 15% of transportation time), saving

millions of dollars in compliance costs, and

effectively curbing tax evasion, laying the foundation

for the modernization of tax management in Brazil.

3.1.2 California Road Tolling Pilot Program

National Grid of the United Kingdom is a

multinational power and gas company headquartered

in London. It is responsible for the power transmission

network in England and Wales and has approximately

20,000 employees in the UK and the United States. Its

revenue in 2023 was approximately 21 billion pounds.

Under the impetus of the EU's "Third Energy

Package" in 2009, the company cooperated with the

government and installed millions of smart meters in

the UK by 2023 to track electricity consumption data

and assist tax authorities in calculating energy taxes.

Pacific Gas and Electric Company (PG&E) is the

largest utility company in California, headquartered in

San Francisco, serving approximately 16 million

customers. Its revenue in 2023 was approximately 24

billion US dollars, and it had about 26,000 employees.

As part of the US federal funding program in 2009,

PG&E deployed over 5 million smart meters in

California, using IoT technology to record energy

consumption and support precise calculation of

electricity bills and related taxes. The background for

these companies to adopt smart meters is the

inefficiency of traditional manual meter reading and

the need for a more flexible billing system in the

context of the rise of renewable energy. As of 2023,

the total number of smart meters worldwide exceeded

1.06 billion, and utility companies ensured efficient

operation of the equipment through cooperation with

technology suppliers such as Siemens and Lansi,

promoting the modernization of tax management.

The California Road Fee Pilot Program is an

innovative measure implemented by the California

Department of Transportation (Caltrans) to address

the issue of insufficient fuel tax revenue. This program

utilizes telemetry technology to collect real-time

vehicle data. The Department collaborates with

technology enterprises to install OBD-II sensors or

deploy smartphone applications on pilot vehicles to

continuously record mileage and transmit the data

wirelessly to the cloud for calculation of

corresponding road fees. To ensure broad applicability,

the system also provides an option for manual input of

mileage. The pilot project, which began in August

2024, covers the entire state of California. Participants

pay fees at the rate of 1.8 cents per mile and receive a

fuel tax credit (Chandra et al, 2020). The assessment

results indicate that this technical solution provides a

feasible alternative for Caltrans under the background

of the reduction in fuel taxes due to the popularization

of electric vehicles. Preliminary data show that the

accuracy of automatic tracking exceeds 95%, and

approximately 90% of the participants pay their fees

on time, confirming that this technology can

effectively support the mileage-based charging model.

The system also collected data on the differences in

road usage between urban and rural areas, providing a

basis for future policy adjustments. Although the pilot

program will continue until January 2025, early

feedback indicates that this measure is expected to

increase road maintenance funds for Caltrans by

several hundred million dollars annually. Caltrans has

pioneered a new way of road taxation through

telemetry technology, verified the feasibility of the

concept, and provided new possibilities for tax

management innovation.

3.2 Specific Applications and Analysis

of AI

3.2.1 Thomson Reuters

Thomson Reuters Corporation was established in

2008 and is a multinational information service giant

Discussion on the Application of Internet of Things and Artiﬁcial Intelligence in Tax Administration

753

formed by the merger of The Thomson Corporation

and Reuters Group. Its headquarters is located in

Toronto, Canada. The Thomson Corporation

originated from the newspaper group founded by Roy

Thomson in 1934 and later expanded into the

multimedia field; Reuters was established by Paul

Julius Reuter in 1851 and has held a leading position

in the global news and financial information services

sector. The strategic integration of the two institutions

combined Thomson's advantages in the North

American market with Reuters' extensive influence in

Europe and emerging markets, forming an important

force in the global information service industry. The

company's business initially focused on news release

and financial data analysis, and then expanded to

professional service areas such as law, taxation,

accounting, risk management, and compliance. Its tax

and accounting business have developed into the

company's core pillar, building a complete product

ecosystem, and serving various accounting firms, tax

departments of multinational enterprises, and tax

regulatory agencies of various countries around the

world. As a knowledge-based enterprise, Thomson

Reuters has accumulated rich technical reserves,

especially the research in the field of AI, which can

be traced back to 1991 when Howard Turtle, the chief

scientist, initiated the AI research project, laying the

foundation for the company's technological

innovation in the following decades.

The global tax environment is becoming

increasingly complex, with frequent changes in tax

laws across different countries, causing tax

professionals to face the dual pressures of information

overload and knowledge update. At the same time,

the compliance requirements for multinational

enterprises' taxes are constantly rising, and accurately

understanding and applying the tax policies in

multiple jurisdictions has become a key challenge. In

the era of digital economy, new business models and

transaction forms have exceeded the applicable scope

of traditional tax frameworks, requiring more

intelligent solutions. Moreover, enterprises generally

hope to improve tax efficiency and reduce

compliance costs. Based on these backgrounds,

Thomson Reuters has developed the Checkpoint

Edge with CoCounsel system, which is a tax research

and compliance tool integrated with generative AI.

This system can understand complex tax inquiries,

extract relevant information from massive tax laws,

regulations, precedents, and professional literature,

and generate structured answers, shortening the

traditional query process from several hours to

several minutes and improving the accuracy of tax

decisions through intelligent analysis. This

technology application reflects Thomson Reuters'

strategy of maintaining its leadership position in the

professional information services sector through

digital transformation, as well as the company's

proactive response to the changing needs of its

customers. By applying AI to automate the

processing of cumbersome tasks and in-depth

analysis of regulatory provisions, Thomson Reuters

has successfully transformed technological

innovation into service value and strengthened its

authoritative position in the global tax information

market.

3.2.2 Intuit TurboTax

Intuit was founded in 1983 and its headquarters is

located in Mountain View, California. It is a pioneer

in the field of financial management software. Its

flagship product, TurboTax, holds approximately

65% of the market share in the personal tax filing

market in the United States, serving over 40 million

users annually. To cope with complex tax laws and

market competition, Intuit has introduced AI

technology into TurboTax since 2018, launching an

intelligent tax-filing assistant system. This system

enables intelligent document recognition and data

extraction (reducing the manual input time from 30-

45 minutes to several minutes), personalized

deduction item recognition (increasing tax refunds by

an average of about $3,900 for users), natural

language interaction interface (lowering the

professional threshold), and anomaly detection

(reducing audit risk by about 40%). These

innovations have reduced the error rate of tax filings

by approximately 65%, shortened the average filing

time from 3.2 hours to 1.5 hours, increased user

satisfaction by 28 percentage points, and reduced

customer acquisition costs by about 22%, thereby

significantly enhancing user experience and market

performance, consolidating Intuit's leadership

position in the tax filing market, and demonstrating

the perfect combination of technology and demand.

4 FUTURE TECHNOLOGICAL

TRENDS AND IMPROVEMENT

SUGGESTIONS

4.1 Future Technological Trends

The combination of IoT and AI shows great potential

in the field of tax management, especially in

ICEML 2025 - International Conference on E-commerce and Modern Logistics

754

intelligent taxation and the application of blockchain

technology. This combination has promoted the

intelligent upgrade of tax management systems,

enhanced compliance, reduced fraud, and improved

efficiency. In the future, blockchain technology will

enhance the transparency and security of tax data,

further optimizing management. With the support of

big data and internet technology, AI has been widely

applied in tax services, and will continue to drive the

scientific, refined, and diversified tax management in

the future. However, current tax services still face

challenges such as technological backwardness,

information asymmetry, talent shortage, and

insufficient data management. Solving these

problems requires government support, encouraging

service institutions to upgrade, strengthening data

sharing and cross-departmental collaboration, and

building an intelligent tax service ecosystem. The

development of automated tax declaration systems

will simplify processes. The combination of AI and

IoT will make declarations more accurate and

efficient, reduce errors, and enhance compliance and

management efficiency.

4.2 Policy Suggestion

4.2.1 The Government Formulates

Reasonable AI + IoT Tax Supervision

Policies

When formulating AI+IoT tax supervision policies,

the government should attach importance to the

promoting role of technological innovation and

enhance the intelligence level of tax source

monitoring and tax enforcement. In terms of tax

source monitoring, strong AI can enhance anti-fraud

capabilities through big data analysis and effectively

combat tax evasion. For instance, the United States

uses individual tax identification numbers to cross-

analyze taxpayer data and precisely identify high-risk

individuals; France summarizes fraud patterns

through "tax document models" and accurately

detects potential risks, demonstrating the

effectiveness of AI in tax risk prevention and control.

In terms of tax enforcement, Brazil has developed an

AI system that automates the processing of repetitive

tasks such as asset search, improving efficiency. Data

from 2022 shows that 35% of judicial cases in Brazil

are tax enforcement cases, and due to insufficient

resources, there are often delays. The introduction of

AI is expected to alleviate this problem and enhance

judicial efficiency.

4.2.2 Balancing the Application of

Enterprise Technologies and Data

Compliance Requirements

With the increasing popularity of AI and IoT

technologies in tax management, enterprises must pay

attention to data compliance requirements while

promoting the application of these technologies.

Enterprises should further improve their own industry

and finance data standard systems and build and

upgrade relevant systems to meet the needs of internal

control and auditing. Specifically, enterprises need to

coordinate the upgrade of systems, implement digital

and intelligent transformation, and achieve integrated

construction of industry and finance to cope with

increasingly complex internal control and external

auditing requirements. Through the introduction of

intelligent technologies, enterprises can enhance the

efficiency of tax management while ensuring data

compliance and avoiding legal and policy risks.

5 CONCLUSIONS

This study explores the application and synergy of

IoT and AI technologies in tax management. With the

rapid development of the global economy and

technology, tax management is confronted with many

challenges, such as data lag, high costs of manual

review, and information silos. The combination of

IoT and AI has brought intelligent changes to tax

management, especially in real-time data collection,

intelligent analysis, and tax risk management,

demonstrating great potential. IoT provides real-time

and precise data collection, significantly enhancing

the timeliness and accuracy of tax management. AI,

through in-depth analysis and intelligent decision-

making of these data, further optimizes tax risk

prediction and management, reducing manual

intervention and human errors. The synergy with AI

not only improves the accuracy of tax monitoring and

VAT management but also promotes the realization

of intelligent tax decision-making.

This study still has some shortcomings in research

perspective, research methods, and research scope. In

the future, this research will improve the research

methods and incorporate more cases to conduct more

in-depth discussions. Although the introduction of

technology has greatly promoted the intelligence and

digital transformation of tax management, there are

still certain challenges in practical application. Issues

such as technology costs, data privacy protection, and

the technical adaptability of tax personnel need to be

gradually resolved on the basis of government policy

Discussion on the Application of Internet of Things and Artiﬁcial Intelligence in Tax Administration

755

support and the continuous maturation of technology.

In the future, with the continuous development of IoT

and AI technologies, tax management will further

move towards intelligence and automation. To

achieve this goal, tax departments need to strengthen

technology application while paying attention to data

compliance and security issues, and promote cross-

departmental collaboration and information sharing

to create a more efficient and transparent tax

management ecosystem.

REFERENCES

Almahri, F. A. A. J., & Saleh, N. I. M. (2025). Insights into

Technology Acceptance: A Concise Review of Key

Theories and Models. Innovative and Intelligent Digital

Technologies; Towards an Increased Efficiency: 2,

797-807.

Chandra, S., Naik, R. T., Venkatesh, M., & Mudgal, A.

(2021). Accessibility evaluations of the proposed road

user charge (RUC) program in California. Transport

Policy, 113, 12-26.

Chen, F. H. (2024). The Breakthrough Path of Power

Material Inventory Management under the Background

of Internet of Things. Modern work teams, (04):35-37.

Merola, R. (2022). Inclusive growth in the era of

automation and AI: How can taxation help?. Frontiers

in artificial intelligence, 5, 867832.

Panhans, M. T. (2024). The rise, fall, and legacy of the

structure-conduct-performance paradigm. Journal of

the History of Economic Thought, 46(3), 337-357.

Shi, Y. N., Song, W., Fan X. L., et al. (2024). Design and

Research on the Data Acquisition and Transmission

System for Tourist Vehicle Supervision Based on

Internet of Things. Internet of Things,14(06):66-69.

Song, M. (2023). Under the new circumstances, tax

administration is confronted with multiple challenges.

Chinese business community, (11):194-195.

Wu, S. T. (2024). A Brief Discussion on Digital

Empowerment of the Construction of New Tax

Administration Capability System. Marketing of Time-

honored Brands, (22):67-69.

Yan, Y. Y. (2021). Real-time Tracking System for Logistics

Information of Storage and Transportation Based on

Internet of Things Technology. China's Warehousing

and Logistics Industry, (09):72-73.

Zheng, S., Trott, A., Srinivasa, S., Parkes, D. C., & Socher,

R. (2022). The AI Economist: Taxation policy design

via two-level deep multiagent reinforcement learning.

Science advances, 8(18), eabk2607.

ICEML 2025 - International Conference on E-commerce and Modern Logistics

756