Artificial Intelligence (AI) in Tax and Customs Trade Compliance

Uncertainty Surrounding the Use of Artificial Intelligence (AI) in Tax and Customs / Foreign Trade Compliance

By Brenda Cordova, Braumiller Law Group Mexico Legal Counsel in collaboration with Saul Rodriguez Corona,
Mexico Attorney, Tax and AI Law Expert

The so-called Fifth Industrial Revolution is no longer a distant concept; it is already shaping everyday life. The term artificial intelligence dates back to 1956, when computer scientist John McCarthy, along with a group of information systems experts, introduced it at the Dartmouth Conference.^[1] Artificial intelligence (AI) has become a common term in public discourse, even as many people remain reluctant to understand it. Yet it is increasingly clear that familiarity with AI will soon be less a matter of curiosity and more a matter of necessity.

At its core, AI refers to the ability to design machines capable of mimicking human intelligence: learning from data, making autonomous decisions, and solving complex problems. While this type of technology feels new, the idea is ancient. Humanity has long imagined tools that could extend or replace human effort, from the mythological automaton Talos in ancient Greece to the work of thinkers such as Descartes, Leibniz, and Alan Turing. AI represents the latest chapter in a long historical pursuit.

So why the sudden surge of attention? The turning point came in June 2020, when OpenAI released GPT-3, a language model that dramatically advanced natural language processing and content generation.^[2] Today, tools such as ChatGPT3 and Gemini demonstrate how AI systems can analyze information and generate responses that closely resemble human communication.

AI is now deeply embedded in the internet and social media, producing everything from humorous content and realistic simulations to advanced automation. These capabilities have expanded beyond entertainment and productivity, reaching into areas of public administration and government oversight.

This raises a critical question: if AI is transforming nearly every sector, how will it reshape tax and customs compliance, and how will it be regulated?

The United States, China, and the European Union have taken notably different approaches. In the U.S., AI regulation is fragmented, spread across more than 120 federal and state laws addressing data protection, technology use, and algorithmic accountability. One notable example is the Take It Down Act, aimed at combating deepfakes and extortion, particularly when minors are involved.^[3]

China, by contrast, has adopted a more centralized model. Its Interim Measures for the Management of Generative Artificial Intelligence Services (2023) impose strict obligations on AI providers, requiring alignment with socialist values and prohibiting content that threatens national security.^[4]

The European Union issued its Artificial Intelligence Act in 2024, establishing a comprehensive framework grounded in the protection of fundamental rights. The EU approach places individuals at the center of AI governance, emphasizing transparency, accountability, and risk prevention.^[5]

Mexico currently lacks binding AI-specific legislation, though several proposals are under consideration. Among them is the Federal Law for the Ethical, Sovereign, and Inclusive Development of Artificial Intelligence (Ley Federal para el Desarrollo Ético, Soberano e Inclusivo de la Inteligencia Artificial) introduced in April 2025, which reflects significant influence from the European regulatory model.

Despite regulatory gaps, governments are not waiting to deploy AI. Tax and Customs authorities have been early adopters, using AI-powered tools for virtual assistance, data analysis, electronic identification, and automated notifications, replacing traditional, in-person processes in the name of efficiency.

Spain’s Tax Agency, for example, has formally committed to using AI to improve taxpayer services, strengthen audits, and promote voluntary compliance. In the United States, the IRS has reportedly implemented AI systems to cross-check data, conduct predictive risk analysis, and flag potential noncompliance.^[6]

In Mexico, artificial intelligence is already reshaping tax and Customs procedures. Through automated risk-analysis systems, Customs authorities can process large volumes of data in real time to detect atypical patterns associated with violations to the law such as undervaluation, incorrect tariff classification, incorrect origin, or potential tax-evasion schemes, among others.^[7] Recent amendments to Mexico’s Customs^[8] and Tax Law, effective in 2026, reinforce this shift toward a more digitalized and data-driven Customs framework. These reforms strengthen the authority’s ability to rely on advanced technological tools for cargo traceability, monitoring of trade operations, and post-clearance audits, reducing dependence on purely manual inspections. While AI promises faster clearance for low-risk operations, it also places taxpayers, importers, exporters, and customs brokers under heightened scrutiny, making data accuracy, consistency, and traceability critical to mitigating legal and tax exposure.

However, this rapid adoption raises serious concerns. Can taxpayers be assured that AI-driven enforcement will not exceed legal limits or undermine due process?

The risks are real. Algorithms may reflect design biases and incorrectly label legitimate transactions as unlawful. Past cases, such as the use of the COMPAS algorithm in U.S. courts, which exhibited racial bias, underscore the dangers of relying on opaque automated systems in high-stakes decisions.

In Mexico, tax and Customs audits fall under discretionary authority, meaning officials are not required to explain why a taxpayer is selected for review. Yet if such decisions are increasingly driven by algorithms, transparency becomes a pressing legal and ethical issue.

The central challenge ahead is ensuring that AI-enhanced tax enforcement complies with fundamental legal principles. While the use of AI by tax authorities is inevitable, failure to establish clear legal safeguards risks creating uncertainty, not only for taxpayers, but for the broader economy.

All eyes now turn to the legislative developments of 2026.

Disclaimer

This article is provided for general informational purposes only and does not constitute legal, tax, customs, or professional advice. The content is based on the legal and regulatory framework in effect as of the date of publication and may be subject to change due to legislative, regulatory, or administrative developments.

The information contained herein is not intended to address the specific circumstances of any particular company or transaction. Tax, foreign trade, Customs, IMMEX, and contractual matters are highly fact-specific, and the application of the law may vary depending on the structure and substance of each operation.

Readers should not act or refrain from acting based on this article without seeking independent professional advice tailored to their specific situation. Neither the authors nor the publisher assumes any responsibility or liability for actions taken based on the information contained in this article.

^[1]PICKOVER Clifford A., Inteligencia Artificial, una visión cronológica, Librero, Madrid, España, 2025, p.93.
^[2]TUSET VARELA, Damián, Marco regulatorio de la Inteligencia Artificial: Derecho Internacional, Unión Europea y Modelos Comparados.

^[3]Take a Down Act, en: https://www.congress.gov/119/plaws/publ12/PLAW-119publ12.pdf, consulta el 22 de enero de 2026.

^[4]Interium Mesures for Generative Artificial Intelligence Service Management, en https://www.sidley.com/-/media/resource-pages/ai-monitor/laws-and-regulations/20230713-china-interim-measures-for-the-management-of-generative-artificial-intelligence-services.pdf?la=en, consulta el 22 de enero de 2026.

^[5]REGLAMENTO (UE) 2024/1689 DEL PARLAMENTO EUROPEO Y EL CONSEJO, en: https://www.boe.es/doue/2024/1689/L00001-00144.pdf, consulta el 22 de enero de 2026.