Introduction
With the rapid spread of generative and agentic AI technologies, fundamental questions have emerged regarding their integration into one of the most sensitive sectors: the judiciary. Justice cannot afford risks to confidentiality, impartiality, or public trust. In this context, the State of California has taken a remarkable step by adopting a new rule that comes into effect on September 1, 2025, to regulate the relationship between courts and this technology.
A Committee and a Clear Timeline
In May 2024, California Chief Justice Patricia Guerrero formed a specialized committee to study the impact of artificial intelligence on the judiciary. This committee worked on developing practical recommendations and, in February 2025, submitted a draft model policy for which courts were asked to provide feedback by April 17, 2025.
After reviewing the feedback, the Judicial Council adopted Court Rule 10.430 and Administrative Standard 10.80, with implementation scheduled for September 1, 2025. Courts were granted a grace period until December 15, 2025, to submit written policies if they chose to adopt artificial intelligence instead of prohibiting it.
Features of the Rule
The rule did not come as general provisions but rather included clear pillars:
- Protecting the confidentiality of judicial data and preventing its leakage to public platforms.
- Requiring human review before approving any technical outputs.
- Disclosure when relying on AI in preparing judicial documents or memoranda.
- Addressing any practices that lead to unlawful bias or discrimination.
- Establishing flexibility that allows courts to align policies with their needs without compromising fundamental principles.
The Saudi Experience
In Saudi Arabia, the judiciary is witnessing a wide-scale digital transformation under Vision 2030. One of its most notable features is the Najiz platform, which has become the unified gateway for digital judicial services, enabling litigants to file lawsuits, follow up on them, and complete electronic notarization efficiently and transparently.
In this context, the regulatory authority has issued the AI Principles and the Generative AI Principles for government entities, both of which stipulate the mandatory application of their provisions across all stages of the lifecycle of these tools to harness their benefits and mitigate risks. These principles were designed to apply to all AI tools across different sectors.
Moreover, the Personal Data Protection Law (PDPL), overseen by SDAIA, serves as a legislative foundation for data protection and strengthens trust in handling sensitive information. This legal and technical infrastructure positions the Kingdom at the forefront of exploring mechanisms for integrating AI into the judiciary in the future, within a framework that preserves justice and reliability.
International Comparison
- European Union: The AI Act classified judicial applications as high-risk categories, imposing the highest levels of oversight.
- United Kingdom and Canada: Both opted for a path of experimentation and gradual testing within legal contexts.
- California: Took a direct practical approach, requiring courts themselves to adopt binding written policies, a step that goes beyond general recommendations.
Conclusion
The decision coming into force on September 1, 2025, reflects a growing conviction that justice cannot remain neutral toward artificial intelligence. The existence of binding written policies ensures that technology becomes an assisting tool, not a source of confusion or bias.
And the message extends beyond California’s borders: every judicial system today faces a similar challenge—between benefiting from digital efficiency and preserving core values. In Saudi Arabia, where judicial digital transformation is led by the Najiz platform and the PDPL, the path seems paved for a new phase that can combine innovation with trust.
Integrating AI into the judiciary is not a matter of “if it will happen” but rather “how it will happen”—and the answer lies in drafting clear frameworks that protect case confidentiality and litigant privacy, ensure human review, and keep justice at the heart of any technological transformation.
You Might Also Like
