When we published AI in the Dock: Africa’s Judiciary Charts a Digital Future, we argued that the continent’s courts are stepping into a defining era, one where technology doesn’t just support justice but also tests its very foundations. Kenya may have just had its first real test.
The Case That Sparked a Larger Debate
In Civil Appeal No. 610 of 2019, Kenya Agricultural and Processed Food Products Development Authority (APEDA) v. Krish Commodities Limited, the Court of Appeal at Nairobi confronted a seemingly straightforward question: could the term “Basmati” be registered as a trademark in Kenya, or does it qualify as a Geographical Indication (GI) reserved for origin-based authenticity?
Yet, amid the legal reasoning, an anomaly appeared. On page 8, paragraph38 of the judgment, the court cited a “Geographical Indications Act, 2019”, a law that does not exist in Kenya’s statute books.
Historically, Kenya’s protection of GIs has been anchored in Section 40A(5) of the Trade Marks Act (Cap. 506), allowing for geographical names to be registered as collective or certification marks. While several drafts for a standalone GI law have circulated, none was enacted in 2019.
So, how did this fictional statute make it into a Court of Appeal ruling?
An AI Hallucination, or a Human Oversight?
Legal analysts suggest this may be a case of AI hallucination, a phenomenon where generative AI systems confidently produce information that sounds credible but is factually incorrect. If confirmed, this could be Kenya’s first public instance of an AI-generated error surfacing in a judicial document.
The reference may have stemmed from an AI model misreading international context or draft legislation or it could be a human oversight amplified by automated drafting assistance. Either way, it exposes a timely truth: AI tools are already in our most human institutions, and we must learn how to use them wisely.
From Alarm to Advancement
It’s tempting to view this as a technological embarrassment. But that would miss the bigger opportunity. As we wrote in AI in the Dock, Africa’s judiciaries are on a journey toward digital maturity, not perfection.
This incident, call it an “AI flub” or a “digital slip”, should not be seen as a crisis, but as a turning point. It proves that AI has quietly entered judicial workflows and now demands governance frameworks as robust as the ones we apply to human ethics.
What the Judiciary Can Learn
- Human Oversight Must Remain the Final Word
AI can accelerate research and drafting, but it must never bypass human review. A second layer of fact-checking and citation validation could prevent such errors before judgments are finalized. - Transparency Builds Institutional Trust
Acknowledging the use of AI and setting clear ethical guidelines for it can reassure the public that automation serves justice, not the other way around. Kenya could pioneer a judicial policy framework on responsible AI use. - Turn Errors into Education
Instead of downplaying the anomaly, the Judiciary could treat it as a case study to strengthen institutional awareness. A small, controlled misstep now can prevent a larger systemic failure later.
Kenya’s Chance to Lead
If Kenya’s judiciary confronts this incident openly, perhaps through a corrigendum and a reflective policy brief, it could set a global standard for judicial integrity in the age of AI.
In most countries, AI-related errors are met with silence or denial. Kenya could flip that script by modeling accountable transparency, acknowledging the issue, correcting it, and using it to guide future digital governance.
That would place Kenya not in the spotlight of scandal, but at the forefront of responsible innovation in justice.
The Broader African Context
Across Africa, courts are exploring AI for case summarization, document analysis, and backlog management. These tools promise speed, but also risk introducing bias, fabrication, or overreliance on machine judgment.
As AI in the Dock noted, “Africa’s courts stand at the intersection of innovation and justice.” This latest Kenyan example sharpens that intersection, showing that the future of fair trials and sound rulings will depend not just on algorithms, but on ethical human governance behind them.
A Teachable Moment, Not a Technological Failure
If AI indeed wrote a line into a Kenyan appellate judgment, it would be a symbolic moment, not of weakness, but of learning and leadership. Every digital transformation carries its early stumbles. What defines progress is not perfection, but how institutions respond to their own errors.
In that light, Kenya’s “AI legal flub” is not an embarrassment. It’s an invitation for the judiciary, the legal fraternity, and technologists to shape an African model of responsible AI in justice: transparent, accountable, and human-centered.
Because in the end, AI may sit on the bench, but it’s still humanity that must deliver the verdict.