This panel comprises four papers investigating the digitalization of tax administration and audit. Each research deals with a particular area of regulation: financial reporting, taxation, anti-money laundering rules, and auditing.First, the research by Magnus Kristoffersson addresses the ethical and regulatory challenges of using artificial intelligence (AI) to detect corporate fraud. His analysis is based on the premise that AI applications should comply with constitutional principles, i.e. “constitutional AI”. The second research by Yuri Matsubara delves into the challenges of using AI for combatting tax fraud in the Japanese context. Her research addresses the reforms passed by the legislators since 2022 and their outcome, as well as the limitations due to strict restrictions regarding the use of taxpayers’ data. The third paper by Eleonor Kristoffersson examines the legal and regulatory interplay between anti-money laundering obligations, advanced AI tools, and the European Union AI Act. Her analysis identifies specific compliance gaps and potential solutions to enhance legal certainty and accountability. The last paper by Clemence Garcia and Hiroko Inokuma addresses the assurance of digital platforms. They review the risks and challenges of auditing blockchain-based systems, namely their decentralization and cryptographic protection. The research concludes with some specific areas of consideration for further research.
As the world becomes increasingly digitized, the intersection of Artificial Intelligence (AI) and financial data has emerged as a critical area of inquiry. The rapid growth of fintech and the potential benefits of AI in streamlining financial operations have sparked substantial interest in this field (Maple et al., 2023). However, this shift toward digitalization also introduces new ethical and regulatory challenges that demand attention. One primary concern involves the impact of AI on privacy and individual rights (Han et al., 2023). The immense volume of financial data processed by AI systems raises significant questions related to data protection, as well as the potential for misuse or exploitation of such data. As the literature emphasizes, financial institutions must implement robust governance frameworks and maintain human oversight throughout decision-making processes to mitigate these risks (Lee, 2020). Another pressing issue is fairness and bias in AI-driven financial services. Without proper design and monitoring, AI algorithms may perpetuate or amplify existing biases, resulting in discriminatory outcomes. Enhancing transparency and accountability in AI decision-making is crucial to ensuring equitable treatment for all individuals. Beyond ethical considerations, the use of AI in financial services also carries substantial policy implications. Regulatory authorities must adopt a proactive approach, moving beyond reactive measures to anticipate and address the systemic risks introduced by AI (Kaneko, 2020; Han et al., 2023; Maple et al., 2023; Lee, 2020). Policymakers should strive to foster a regulatory environment that encourages innovation while safeguarding consumer rights and interests. Constitutional AI is an emerging concept that seeks to align AI systems with core constitutional principles such as human rights, democracy, and the rule of law (Abiri, 2024). Integrating these principles into the design and implementation of AI in financial services can help mitigate the aforementioned risks and challenges. This paper will examine the general concept of Constitutional AI in relation to the use of digitized financial data and AI-driven tools. Although the study is grounded in the field of legal science, it also intersects with computer science and management perspectives, including stakeholder theory, to provide a holistic understanding of the relevance and impact of AI on law and corporate governance (Study on the Relevance and Impact of Artificial Intelligence for Company Law and Corporate Governance, 2021).
Currently, the Japanese government is accelerating digitalization. This article discusses how taxpayers adapt to the new regulations, especially regarding Tax and AI. It also discusses transfer pricing as a case of using AI in taxation. This article investigates how tax administrations and practitioners use AI for tax compliance. On the administration’s side, AI is used to detect tax fraud. In Japan, AI is understood to streamline normal administrative procedures, and the Japanese government did not decide to restrict its use (especially generative AI -after 4.5-). An AI application, the AI chat pot, was also introduced in 2022. According to the press release on 30 Oct 2024, the Japanese tax authority gained a historical record amount to collect “unpaid corporate income taxes” as a consequence of using AI. Concretely, it is used to replace skilled tax inspectors.
This paper examines the legal and regulatory interplay between anti-money laundering (AML) obligations, advanced artificial intelligence (AI) tools, and the evolving European Union AI Act (See Kingdom 2004). Building upon a critical analysis of the EU AI Act, the study explores how current and emerging AI-driven AML solutions can effectively detect illicit financial activities while upholding fundamental privacy and fairness principles (Radanliev et al., 2024). It investigates the tension between the need for transparency and interpretability of automated decision-making, the proportional use of personal data, and the challenges posed by algorithmic opacity (Felzmann et al., 2020). In doing so, the paper identifies specific compliance gaps and outlines potential pathways for ensuring that AML initiatives not only align with the EU law but also foster legal certainty, accountable innovation, and sustained public trust (Koene et al., 2019).
Beyond bitcoin and fintech, various chains serve operational purposes like logistics, managing sensitive data, and even corporate governance. In this context, our research addresses the audit procedures for these decentralized platforms. Since blockchain is often associated with fraud, tax evasion, and money laundering, it is worth considering to what extent assurance can be provided on these systems. The challenges of auditing information systems have been extensively discussed in academic and professional literature. This article builds on prior research (Bedard, 2005; Sheldon, 2019; Yang et al, 2020) to analyze the specific challenges and risks regarding blockchain-based information systems. After a review of the state-of-the-art audit practices in the area, we delve into specific issues like managing keys and access tokens, the security of wallets, transaction signing, etc. Last, we review the overall assessment of IT controls and governance.