Extended Abstract
Background and Purpose
The expanding use of artificial intelligence in professional sports—from talent identification to refereeing and competition management—has increased efficiency but also created new legal challenges. One major challenge is "algorithmic discrimination," which arises from biased training data, opaque system design, and lack of oversight, potentially reproducing or exacerbating inequalities against certain athletic groups. Despite global regulatory efforts (e.g., EU), a significant gap exists in the Iranian legal system: although constitutional principles prohibit discrimination, no specific regulations address AI governance or algorithmic discrimination in sports. Therefore, this study aims to analyze the legal dimensions of algorithmic discrimination in sports and propose regulatory solutions, including a multi-level framework tailored to Iran.
Methods
This research adopts a fundamental theoretical and descriptive-analytical methodology, employing doctrinal legal analysis as its primary method, which focuses on the systematic interpretation of laws, regulations, international instruments, and legal doctrines. Given the complex and emerging nature of algorithmic discrimination in sports, this approach enables an in-depth legal investigation without conducting field or statistical studies. Within this framework, four complementary techniques are utilized: (1) structured literature review for targeted source collection; (2) conceptual analysis to define key notions such as algorithm, algorithmic decision-making, and algorithmic discrimination; (3) systematic comparative analysis to examine selected legal approaches; and (4) legal reasoning to interpret existing rules, identify normative gaps, and propose corrective solutions. The documentary population includes national and international laws, policy documents, reports from international bodies, ethical charters, and scientific articles published between 2015 and 2025. Sources were searched in Google Scholar, Scopus, SSRN, and Iranian databases (Noormags, Magiran, Elmnet, Civilica). After screening 36 initial sources, 26 were finally selected: 13 peer-reviewed articles/books, 8 legal instruments (including constitutions, the EU AI Act, Iranian national laws, and international conventions), 3 regulatory reports (EEOC, FTC, NIST), and 2 ethical frameworks (UNESCO 2021, IOC 2024). Comparative analysis followed a four-stage systematic process: extracting comparison criteria (legislation status, legal bases, scope, technical requirements, transparency, enforcement); selecting four units (EU, US, international sports bodies, and Iran); collecting and categorizing equivalent data; and critically evaluating strengths and weaknesses. Data analysis was qualitative, using conceptual analysis and inductive legal reasoning to interpret constitutional principles (especially Articles 3, 19, and 22 of the Iranian Constitution) and to provide regulatory proposals. Despite limitations such as lack of empirical data, rapid technological change, and restricted access to proprietary algorithmic systems used in Iranian professional sports, this research provides a theoretical and legal foundation for future policy-making on AI governance and the prevention of algorithmic discrimination in Iran.
Results
The findings of this research reveal that algorithmic discrimination in professional sports manifests in multiple distinct forms, each with specific legal and ethical implications. First, sampling bias occurs when AI models are trained predominantly on data from male elite athletes, leading to inaccurate predictions and systematic disadvantages for women, youth, and athletes with disabilities. Second, labeling bias emerges from inconsistent definitions of concepts such as injury across different sports contexts or between wealthy and poor clubs, resulting in unequal access to medical care and support. Third, feature bias arises when algorithms focus exclusively on easily measurable physical data while neglecting critical psychosocial factors like stress or mental pressure, creating an unfair care system that marginalizes athletes' holistic needs.
Fourth, automation bias involves over-reliance on AI outputs without meaningful human review, potentially depriving athletes with unusual profiles of fair, personalized evaluation by human experts. Fifth, commercial and structural bias stems from proprietary "black box" models used by wealthy clubs, creating unfair competitive advantages and prioritizing financial interests over athlete welfare. These manifestations collectively violate the fundamental sporting principles of competitive justice and equal opportunity as enshrined in the Olympic Charter and international human rights instruments.
The analysis identifies four interrelated root causes of algorithmic discrimination. First, bias in training data reflects and perpetuates historical societal inequalities, with underrepresentation of minority groups leading to systematically biased outcomes. Second, algorithm design and architecture often operate as opaque systems lacking fairness-aware mechanisms, with proxy variables such as height, weight, or geographic location inadvertently institutionalizing structural inequalities. Third, institutional factors and regulatory gaps manifest in weak oversight and the absence of specific legal frameworks for AI in sports, with organizations prioritizing economic efficiency over equity considerations. Fourth, human and cultural factors include the prioritization of optimization and profitability over fairness, coupled with the absence of regular auditing and continuous evaluation.
The comparative analysis of international approaches yields significant results. The European Union, through its Artificial Intelligence Act (2024), has established a comprehensive, risk-based regulatory framework with binding obligations for high-risk AI systems, including those used in sports. Requirements include data quality assurance, transparency, explainability, human oversight, and conformity assessments, backed by substantial fines of up to 35 million euros or 7% of global turnover. This represents the world's first comprehensive hard law approach to AI governance.
In contrast, the United States lacks a unified federal AI law, instead applying existing anti-discrimination legislation such as Title VII of the Civil Rights Act of 1964 to algorithmic tools. The Equal Employment Opportunity Commission and Federal Trade Commission have issued guidance holding employers fully responsible for AI-induced discrimination, applying the "disparate impact" doctrine to seemingly neutral algorithms that produce disproportionately adverse effects on protected groups. The National Institute of Standards and Technology has developed voluntary risk management frameworks, though these lack legal enforcement mechanisms.
International sports bodies have adopted ethical and reputational approaches. The International Olympic Committee's 2024 Framework on AI in Sport emphasizes transparency, fairness, accountability, and human rights. The World Anti-Doping Agency's 2025 Code mandates meaningful human oversight, prohibiting exclusively algorithmic decisions in doping cases. UNESCO's 2021 Recommendation on the Ethics of AI establishes equality and non-discrimination as fundamental principles, providing a common ethical foundation adopted by over 190 member states.
Regarding the Iranian legal system, the research reveals significant findings. The Constitution provides general foundations for opposing discrimination through Article 3 (elimination of unjust discrimination), Article 19 (equal rights regardless of race, color, language), and Article 22 (protection of livelihood and dignity). International commitments, including Iran's accession to the International Convention against Apartheid in Sports (1997), establish binding obligations for non-discrimination in sports. However, the research identifies a critical regulatory vacuum: no specific legal framework addresses algorithmic discrimination or AI governance in sports. The Electronic Commerce Act (2004), the only existing law related to data protection, is fundamentally inadequate for addressing algorithmic discrimination due to its limited scope and objectives.
This regulatory gap is particularly problematic because general anti-discrimination provisions, while applicable to public sector actions, lack effectiveness against private sector algorithmic systems where most discrimination occurs. Furthermore, effective prevention requires specific, technically informed regulations rather than reliance on general principles. The absence of specialized oversight bodies, transparency requirements, pre-deployment impact assessments, and clear accountability mechanisms leaves athletes vulnerable to algorithmic discrimination without adequate legal recourse. The findings underscore the urgent necessity for comprehensive, multi-level regulatory intervention tailored to the Iranian context while drawing upon successful international experiences.
Algorithmic discrimination in professional sports represents a profound legal and ethical challenge that transcends technical flaws, reflecting and amplifying historical structural inequalities. This phenomenon undermines competitive justice and equal opportunity—the foundational pillars of sports ethics and international legal commitments. The comparative analysis reveals diverse regulatory approaches: the European Union's binding risk-based framework, America's enforcement of existing anti-discrimination laws, and international sports bodies' ethical guidelines. Within the Iranian legal system, despite constitutional principles prohibiting discrimination, a critical regulatory vacuum exists regarding AI governance in sports. Effective prevention requires a multi-level strategy encompassing legislative action (adopting risk-based regulations), institutional strengthening (empowering oversight bodies), technical standards (ensuring transparency and human oversight), and international engagement. Proactive intervention is not merely an option but an unavoidable necessity to preserve the fair and inclusive spirit of sport in the digital age and protect athletes' fundamental rights against emerging technological threats.
Article Message
This article examines algorithmic discrimination in professional sports within the Iranian legal context. As artificial intelligence increasingly governs talent identification, performance analysis, and refereeing, algorithmic systems risk perpetuating historical biases against women, ethnic minorities, and disabled athletes. Through comparative analysis of EU regulations, US enforcement approaches, and international sports frameworks, this research reveals a critical regulatory vacuum in Iranian law despite constitutional anti-discrimination principles. The article proposes a multi-level governance model encompassing legislative, institutional, technical, and international strategies to ensure responsible AI deployment in sports while protecting athletes' fundamental rights and preserving competitive justice.
Ethical Considerations
This research was conducted in accordance with fundamental ethical principles of academic integrity and honesty. All sources have been properly cited and acknowledged, and the work of other scholars has been respected through accurate attribution. The research avoids any form of plagiarism, data fabrication, or falsification. Given the documentary and library-based nature of this study, no human subjects were involved, thus exempting it from informed consent requirements. However, the research maintains objectivity and impartiality in analyzing legal systems and frameworks, presenting findings without bias toward any particular jurisdiction or approach. The study acknowledges its limitations, including the absence of empirical data and the rapid evolution of AI technologies, and refrains from overgeneralizing its conclusions. Furthermore, the research recognizes the potential societal implications of algorithmic discrimination and emphasizes the importance of protecting vulnerable groups in sports, aligning with the ethical imperative to promote justice, equality, and human dignity in all scholarly endeavors.
Conflict of Interest
There is no conflict of interest in this study.
Acknowledgments
All the loved ones who participated in the present study are appreciated and thanked.