The Idea–Expression Dichotomy in the Copyright Bill, 2026

Copyright law is built on a distinction between ideas and their expression. While ideas remain free for public use, their particular expression is protected to encourage creativity and innovation. This principle, known as the idea–expression dichotomy, lies at the core of modern copyright systems and is reflected in both international instruments and  the recently tabled Copyright Bill, 2026 (Kenya).

However, despite its conceptual clarity, the application of this doctrine has become increasingly complex in light of technological developments, particularly in software development and generative artificial intelligence.

1. The Legal Framework of Copyright Bill

The idea–expression dichotomy is embedded in the recently tabled Kenya Copyright Bill, 2026, under Sections 17 and Section 25, and international standards under Article 2 of the WIPO Copyright Treaty. These provisions affirm two key principles:

• Copyright subsists in original works fixed in a tangible medium of expression; and
• Protection does not extend to ideas, procedures, methods, systems, or concepts.

2. The Problem: The Blurring of Idea and Expression

Notwithstanding its theoretical appeal, the idea-expression distinction is difficult to apply in practice. The boundary between an unprotectable idea and a protectable expression is often thin and unstable.

In some contexts, the constraints of functionality mean that expression is often dictated by the underlying idea itself, leaving limited room for creative variation.

Let’s look at it this way;  a mobile banking application designed to enable users to transfer money, check balances, and pay bills is usually driven by a functional idea: facilitating digital financial transactions. However, because efficiency and usability require standardized solutions, developers often rely on similar code structures, database architectures, and user interface designs. In such cases, separating idea from expression becomes artificial, and copyright protection risks either overextending into functional ideas or failing to adequately protect genuine creative effort.

Secondly, the rise of generative artificial intelligence has further complicated the idea–expression framework. Say a user prompts an AI system to produce content such as “write a legal contract for software licensing” the human input is typically limited to a broad instruction. The AI system then autonomously determines the expressive elements of the output, including wording, structure, and stylistic choices.

In this process, the human user exercises minimal control over the final form of expression. Instead, the technology itself effectively translates an abstract idea into a completed expressive work.

This raises a critical doctrinal question: at what point does copyright protection attach when expressive choices are no longer made by a human author?

Regulatory guidance, including that of the United States Copyright Office, has increasingly emphasized that copyright protection requires human authorship. Where expressive elements are determined autonomously by AI systems, such outputs may fall outside the scope of copyright protection. This reflects an emerging tension within copyright law, as technological systems begin to disrupt traditional assumptions about authorship and creativity.

3. Conclusion and Recommendation

While the Copyright Bill, 2026 & International framework continues to uphold the distinction between idea and expression, its practical application requires greater clarity in areas where expression is heavily constrained or machine-generated. Perhaps as the jurisprudence around this area develops, we might need more refined statutes to determine when expression is truly independent of underlying ideas. This will likely clarify whether the idea–expression dichotomy remains a workable legal standard or evolves into a more nuanced framework for the digital age.