Kenya Strengthens Rules on Plagiarism and AI-Use Thresholds in University Theses

Nairobi, Kenya – In a step to maintain high academic standards, Kenyan universities are adopting clear rules for plagiarism and use of artificial intelligence (AI) in theses. The new thresholds are:

• Less than 10% similarity = Acceptable

• 10–15% similarity = Needs Evaluation

• Above 15.1% similarity = Fail

These rules apply to all academic theses submitted for master’s and doctoral degrees across public and private universities.

Why the New Thresholds?

Academic institutions in Kenya have seen growing concern over plagiarism and misuse of AI tools by students. Some students use AI to generate large portions of text, then submit it without editing, citing, or acknowledging. Others have high similarity scores on plagiarism detection tools. Universities want to ensure originality, academic integrity, and that students develop their own critical thinking and writing skills.

How It Will Be Enforced

• Universities will use plagiarism detection software on all theses.

• Any thesis with a similarity score of 10–15% will be reviewed by a committee. The student may be asked to revise, clarify sources, or explain AI-tool usage.

• Theses above 15.1% similarity will be rejected (fail) immediately, unless there is a very strong justification and evidence of mistakes in detection.

Role of AI

AI tools (like chatbots, large language models) may be used by students, but only if:

• Their contribution is clearly disclosed and cited.

• The student edits, reflects, and adds original work beyond what the AI tool generated.

• The final work shows student’s own ideas, analysis, and writing.

Use of AI without disclosure, or over-dependence on AI (letting it write large parts without student input) will lead to high similarity scores and possible failure.

Benefits and Challenges

Benefits:

• Enhances trust in Kenya’s higher education credentials.

• Encourages students to learn proper research, writing, and critical thinking.

• Helps Kenya compete internationally in academic standards.

Challenges:

• Some universities may lack good plagiarism-detection tools.

• Students and staff need training in how to properly use AI, how to cite AI-generated content, and avoid unintentional plagiarism.

• Defining what counts as “AI-generated” vs student-written content can be tricky.

What Students Must Do

• Always run similarity/plagiarism checks before final submission.

• Keep records of all sources, including AI tool usage (what prompts, what output, how you modified).

• Cite any AI-assisted work.

• Ensure your thesis is mostly your own writing, ideas, and research.