The Judiciary of Kenya has moved to regulate the growing use of artificial intelligence in courtrooms after judges encountered legal filings containing fake AI-generated case citations and factual inaccuracies.
The Judiciary released a Draft Judiciary Artificial Intelligence Policy aimed at establishing rules governing how lawyers, judicial officers and court systems may use AI technologies within Kenya’s justice system.
The move follows a March 2026 incident at the Milimani High Court where a judge struck out an AI-generated legal application after discovering that the filing contained references to non-existent cases and inaccurate legal material.
The incident intensified concerns about the rapid adoption of generative AI tools within the legal profession, where lawyers increasingly use AI for legal drafting, research and case preparation.
Analysts say Kenya is now among the first African countries attempting to formally establish AI governance structures specifically tailored to judicial systems.
Fake AI Citations Triggered Judiciary Response
The Judiciary AI policy emerged partly in response to growing fears surrounding AI “hallucinations” — situations where generative AI systems produce fabricated information presented as factual.
In the Milimani High Court case, the filing reportedly relied heavily on AI-generated content that included false legal citations and inaccurate references.
The court rejected the filing entirely, sending a strong warning to advocates increasingly relying on AI-generated legal drafting without proper verification.
The incident highlighted one of the biggest risks associated with generative AI within legal systems.
Unlike traditional legal databases or research software, large language models can generate convincing but entirely fictional legal authorities if outputs are not carefully verified.
Legal experts say such errors pose serious risks because court decisions depend heavily on factual precision, accurate precedent and procedural integrity.
The Judiciary’s response attempts to balance technological innovation with legal accountability rather than imposing an outright ban on AI use.
Judiciary AI Policy Keeps Humans in Control
One of the central principles within the proposed Judiciary AI policy is the requirement for “human-in-the-loop” oversight.
Under the draft framework, AI systems may assist judicial officers and lawyers in areas such as:
- Legal research
- Scheduling
- Case management
- Transcription
- Administrative support
However, AI tools will not be permitted to replace human judicial decision-making.
The policy explicitly states that judges and judicial officers must remain the final decision-makers in all court processes.
Analysts say this reflects growing international consensus that AI should augment rather than replace judicial reasoning.
The draft policy emphasizes that constitutional responsibility for justice cannot be delegated to automated systems.
Lawyers Must Verify AI-Assisted Filings
The Judiciary AI policy introduces a major accountability measure aimed directly at advocates using generative AI tools.
Any legal filing assisted by AI would require a certificate of human verification confirming that the lawyer has personally reviewed and verified the accuracy of the content.
The rule effectively places legal responsibility on advocates rather than software systems.
The Judiciary argues that lawyers must remain accountable for everything submitted before courts, regardless of whether AI tools contributed to the drafting process.
Legal analysts say the requirement may become one of the most significant practical safeguards within the policy.
It directly addresses concerns that some legal practitioners may increasingly rely on AI-generated research without conducting independent verification.
The policy therefore seeks to preserve professional accountability while still permitting limited AI-assisted efficiency gains.
Judiciary AI Policy Creates Risk-Based AI Categories
The proposed framework also introduces a tiered AI classification system modeled partly on Kenya’s broader Artificial Intelligence Bill, 2026.
Under the draft, AI systems would be categorized according to risk levels and potential impact on judicial outcomes.
High-risk systems would include tools used in areas such as:
- Predictive sentencing
- Bail assessments
- Judicial recommendation systems
Such tools would require mandatory human oversight, periodic audits and heightened scrutiny.
Mid-tier systems include legal research platforms where lawyers must verify outputs and disclose AI usage.
Lower-risk tools such as scheduling systems and transcription software would face lighter monitoring requirements.
Analysts say the tiered approach reflects a more sophisticated regulatory model than blanket restrictions.
Rather than banning AI entirely, the Judiciary AI policy attempts to calibrate oversight according to the seriousness of potential consequences.
Data Protection and Bias Become Major Concerns
The Judiciary AI policy also addresses concerns about privacy, bias and fairness.
Court systems routinely process highly sensitive personal information involving criminal proceedings, family disputes, financial records and constitutional matters.
The draft therefore requires all AI systems used within judicial processes to comply with Kenya’s Data Protection Act, 2019.
The policy also warns about algorithmic bias.
Unchecked AI systems may reproduce discriminatory patterns or unfair outcomes, particularly against already marginalized groups.
Global research has shown that AI systems trained on biased datasets can reinforce inequalities in areas including criminal justice, employment and financial services.
The Kenyan Judiciary says safeguards must therefore ensure that technology does not undermine constitutional principles of fairness and equality before the law.
Digital Access Gap Remains a Challenge
The Judiciary AI policy also recognizes a broader social challenge tied to technological adoption.
Judicial efficiency gains may have limited value if they widen access gaps between digitally literate citizens and those lacking internet access or technical skills.
To address this concern, the Judiciary proposes using ICT support desks at Huduma Kenya centers to help citizens interact with digital judicial systems.
Analysts say digital exclusion remains a major issue across many African legal systems as governments expand online services and electronic filing systems.
Without support structures, vulnerable populations could struggle to access increasingly digitized justice systems.
Kenya Aligns With Global AI Court Guidelines
The Judiciary AI policy also aligns closely with international discussions around AI governance in courts.
The draft references UNESCO’s 2025 guidelines on artificial intelligence in judicial systems, which emphasize transparency, accountability and non-delegation of judicial authority to machines.
However, Kenya’s proposal goes further in certain areas.
One of the most ambitious elements involves plans for a connected digital system linking:
- The Judiciary
- The police service
- The Office of the Director of Public Prosecutions
- The prisons service
Such integration could significantly streamline case management and information sharing across Kenya’s justice system.
Analysts say implementation, cybersecurity and privacy protections will become critical if such interconnected systems are developed.
AI Adoption Expands Across Legal Profession
The Judiciary AI policy reflects how rapidly generative AI has entered legal practice globally.
Law firms increasingly use AI systems for document drafting, legal summarization, contract analysis and research assistance.
Supporters argue that AI can reduce repetitive work, improve efficiency and lower legal service costs.
Critics, however, warn that overreliance on AI may weaken professional standards and increase risks of misinformation or procedural errors.
The Kenyan Judiciary’s approach suggests authorities expect AI adoption to continue expanding rather than disappear.
The policy therefore focuses on controlled integration rather than prohibition.
Why This Matters
The Judiciary AI policy positions Kenya among the first African countries attempting to formally regulate AI use within court systems.
The framework highlights growing concerns about misinformation, accountability and bias as generative AI tools spread rapidly across professional sectors.
The policy may also influence broader discussions about AI governance across Africa’s legal, public sector and regulatory institutions.
What Happens Next
The draft Judiciary AI policy has now entered the public participation phase, where legal professionals, civil society groups and technology experts are expected to submit feedback.
Final implementation will likely depend on how effectively the Judiciary balances technological innovation with constitutional protections, professional accountability and public trust.
For Kenya’s legal sector, the policy could become a defining framework shaping how AI is integrated into courts over the coming decade.
