Kenya's Judiciary AI Policy Signals a New Phase of AI Governance, But Harmonisation Must Follow

As Kenya awaits the enactment of the National Artificial Intelligence Policy, the Judiciary has become the first arm of government to publish a sector-specific framework governing the responsible use of AI. The Draft Judiciary Artificial Intelligence Policy signals how AI governance in Kenya may evolve sector by sector before a national framework comes into force. While this proactive approach demonstrates institutional leadership, it also raises an important question: how should sectoral AI policies interact with an eventual national AI policy?

The Judiciary Steps Ahead of the National Framework

Kenya's National Artificial Intelligence Policy remains in draft form. Yet, even before its adoption, the Judiciary has moved to establish its own governance framework for artificial intelligence.

Developed following extensive research and stakeholder engagement involving the Centre for Intellectual Property and Information Technology (CIPIT) at Strathmore University, the Draft Judiciary Artificial Intelligence Policy seeks to guide the ethical, responsible and accountable use of AI within Kenya's justice system.

The policy comes at a time when AI tools have become increasingly common in legal practice. Lawyers, judges, researchers and litigants are using generative AI to assist with legal research, drafting submissions, summarising authorities and preparing documents. While these technologies promise greater efficiency, they have also exposed new risks, including fabricated authorities, algorithmic bias, confidentiality concerns and diminished professional accountability.

Recent reports of AI-generated pleadings containing fictitious case citations have demonstrated that these risks are no longer theoretical. The Judiciary's decision to develop a dedicated AI policy reflects an acknowledgement that existing professional obligations alone may not sufficiently address the challenges created by rapidly evolving AI technologies.

Rather than waiting for national legislation, the Judiciary has opted to establish principles that govern AI use within its own constitutional mandate.

Why Sectoral AI Policies Matter

The Judiciary's initiative illustrates an increasingly important feature of modern AI governance: Sectoral regulation.

AI does not affect every industry in the same way. The risks associated with AI in healthcare differ fundamentally from those arising in financial services, education, policing or the administration of justice. While national AI policies establish broad principles, sector-specific frameworks translate those principles into operational standards suited to particular institutional contexts.

This approach has become increasingly common internationally.

The European Union's AI Act adopts a risk-based framework while allowing sector-specific regulators to oversee AI applications within their respective mandates. Singapore supplements its national AI governance model with detailed guidance for finance, healthcare and public services through the Model AI Governance Framework and sector-specific initiatives led by the Monetary Authority of Singapore. In the United States, although no comprehensive federal AI law exists, agencies such as the Food and Drug Administration (FDA), Securities and Exchange Commission (SEC) and Equal Employment Opportunity Commission (EEOC) have each issued AI guidance relevant to their respective sectors.

Kenya's Judiciary is therefore following an emerging global trend: recognising that AI governance cannot rely solely on broad national principles but must also respond to the practical realities of individual sectors.

What the Draft Judiciary AI Policy Seeks to Achieve

The draft policy seeks to integrate AI into judicial administration while preserving the core values that underpin the administration of justice.

Its objectives include promoting responsible innovation, improving efficiency in court administration and supporting access to justice while ensuring that AI remains subject to human oversight.

Several themes run consistently throughout the policy.

First, the policy makes clear that AI is intended to assist, not replace, judicial decision-making. Judicial reasoning and the determination of disputes remain exclusively human responsibilities.

Second, it places significant emphasis on transparency and accountability. Users of AI systems are expected to understand both the capabilities and limitations of these technologies rather than unquestioningly relying on automated outputs.

Third, the policy recognises the importance of protecting confidentiality, privacy and data security. Given the sensitive nature of court records and judicial proceedings, these safeguards are essential to maintaining public confidence in the justice system.

Finally, the framework promotes responsible innovation by encouraging experimentation while ensuring that AI deployment remains consistent with constitutional values, judicial independence and the rule of law.

Together, these principles reflect internationally recognised responsible AI standards, including human oversight, fairness, transparency, accountability and respect for fundamental rights.

Benefits Beyond the Judiciary

Although designed specifically for the courts, the policy's significance extends beyond judicial administration.

For legal practitioners, it offers greater clarity regarding the responsible use of AI in litigation and legal drafting. Clear institutional expectations may reduce the likelihood of professional misconduct arising from overreliance on generative AI systems.

For technology developers, the policy provides regulatory certainty. Developers seeking to build AI tools for the justice sector gain clearer guidance regarding acceptable standards, ethical safeguards and governance expectations.

For the public, the framework demonstrates that AI adoption within the justice system will remain subject to constitutional safeguards rather than being driven solely by technological capability.

Perhaps most importantly, the policy establishes an institutional precedent. Other public institutions, including regulators, Parliament, county governments and independent commissions, may now consider developing sector-specific AI governance frameworks suited to their own operational environments.

The Risks of a Sector-by-Sector Approach

While sectoral policies provide flexibility, they also create the possibility of regulatory fragmentation.

If each institution develops AI governance independently, Kenya could eventually have multiple frameworks containing different definitions, risk classifications, compliance obligations and oversight mechanisms. Such inconsistency would create uncertainty for public institutions, technology developers and businesses operating across multiple sectors.

This challenge becomes particularly significant while the National AI Policy remains under development.

Without an overarching framework, individual institutions may adopt standards that differ in important respects, making future harmonisation more difficult.

International experience suggests that successful AI governance depends upon balancing institutional autonomy with national consistency. Sector-specific guidance works best when built upon shared national principles that ensure coherence across government.

The Need for Harmonisation

The Judiciary's Draft AI Policy should therefore be viewed not as a substitute for Kenya's National AI Policy but as an important building block within a broader national governance framework.

Once the National AI Policy is adopted, sector-specific AI frameworks should be reviewed and aligned with national standards to ensure consistency in terminology, governance structures, risk assessment methodologies and accountability mechanisms.

A harmonised framework would preserve the flexibility necessary for individual sectors while avoiding regulatory duplication and conflicting obligations.

Such an approach would also strengthen Kenya's position as it seeks to implement the African Union Continental AI Strategy and participate in broader international AI governance initiatives.

Rather than viewing sectoral policies and national policy as competing instruments, policymakers should recognise them as complementary layers of governance: the national framework providing common principles, while sector-specific policies translate those principles into operational rules for particular institutions.

Conclusion

The Judiciary's Draft Artificial Intelligence Policy represents a significant milestone in Kenya's AI governance journey. It demonstrates that public institutions need not wait for comprehensive national legislation before establishing responsible AI safeguards within their own mandates.

At the same time, the policy highlights the next challenge facing Kenya's AI regulatory landscape.

Sectoral innovation should not evolve into sectoral fragmentation.

As Kenya finalises its National Artificial Intelligence Policy, policymakers should seize the opportunity to create a coherent governance architecture in which sector-specific AI frameworks operate within a common national framework. Such harmonisation would provide regulatory certainty, promote responsible innovation and ensure that artificial intelligence advances public service delivery without compromising constitutional rights, judicial independence or public trust.