Beyond Regulation: The Future of AI in the Courts
The regulation of artificial intelligence (AI) varies significantly across the globe. While some jurisdictions have established comprehensive frameworks, others rely on sector-specific rules, and there are those that resist regulation entirely. A notable example is the European Union’s AI Act, which implements a horizontal, risk-based model applicable across sectors and technologies. In contrast, the United States faces ongoing resistance to binding legislation, with discussions around a proposed moratorium on state AI regulation that could last over a decade.
This discrepancy in regulatory approaches has influenced the global dialogue surrounding AI governance, often focusing on potential legislation and policy design. However, this focus tends to obscure a crucial and often overlooked reality: courts are already regulating AI.
Current Judicial Involvement
Judges, data protection authorities, and administrative tribunals are actively addressing disputes involving AI systems. These disputes range from matters of automated immigration decisions and biometric surveillance to data processing for model training. In this context, courts are establishing de facto rules, boundaries, and precedents that significantly shape how AI technologies are developed, deployed, and challenged.
This study presents findings from an empirical project that tracks and categorizes litigation involving AI technologies across various jurisdictions. The dataset comprises 500 cases from 39 countries, including both judicial decisions and administrative rulings. Each case involves a direct or materially significant use of AI and has been manually classified by jurisdiction, court or administrative authority, AI-related issue, and year.
Key Areas of Litigation
Among the analyzed cases, three issue categories account for 39.4% of all disputes:
- Legal profession (92 cases)
- Intellectual property (56 cases)
- Administrative use of AI (49 cases)
These categories are not only quantitatively dominant but also represent significant friction points between AI systems and the legal frameworks governing them.
Legal Profession Litigation
The rise of litigation within the legal profession is particularly noteworthy. Cases surged from just three in 2021 to 32 in 2024, with an additional 28 already documented in early 2025. These cases involve the use of AI by legal practitioners and its citation in judicial reasoning. An illustrative example is the case of Ross v. United States (2025), where the DC Court of Appeals referenced OpenAI’s ChatGPT to interpret the concept of “common knowledge” in an animal cruelty case. Similarly, in Brazil, a lawyer submitted a petition containing 43 fabricated precedents generated by an AI model, leading to the case’s dismissal and a disciplinary inquiry.
Intellectual Property Disputes
Intellectual property cases have also seen a dramatic increase. From fewer than five cases annually before 2022, this category peaked at 27 cases in 2024, largely driven by the emergence of generative AI. Many of these disputes involve claims of copyright infringement, such as in the case of Getty Images v. Stability AI, where plaintiffs argue that using millions of copyrighted images for model training violates IP laws.
Administrative Use of AI
The use of AI in administrative contexts has been a contested area for a longer time. Cases in this category date back to 2014 but notably increased between 2019 and 2023, reaching a peak of 13 cases in 2023. These cases often concern automation in public services, including immigration and benefits distribution. For example, in Haghshenas v. Canada (2023), a complainant contested a visa denial influenced by an AI tool used for immigration processing, revealing complexities in hybrid decision-making and accountability.
Implications for AI Governance
These litigation trends illustrate not only the areas where legal disputes are emerging but also how judicial systems are becoming key players in AI governance. Courts are not merely responding to harms but actively shaping new legal boundaries for emerging technologies. They define what constitutes fair process or lawful innovation in contexts where prior legal guidance is often ambiguous or lacking.
As society seeks to govern AI in a manner that reflects real-world dynamics rather than abstract principles, it is crucial to recognize the role of the courtroom in this discussion. The law is already articulating its stance on these issues; the challenge now is to ensure that these judicial rulings are integral to the ongoing conversation about the future of technology and democracy.
