REKEL PATENTINTELLECTUAL PROPERTY

IP / CASE STUDY

Google v. Oracle: when using an API can qualify as fair use

Android, the Java interface and the boundaries of software protection: what the US Supreme Court decided and which questions it deliberately left open.

Background: a familiar interface in a new environment

When building Android, Google wrote its own implementing code while using approximately 11,500 lines of declaring code from the Java SE interface. The purpose was to allow programmers to use familiar commands in a mobile environment. Oracle, the rights holder in Java SE, challenged that use. The dispute therefore concerned a particular layer of software rather than the straightforward appropriation of an entire system. [2]

The procedural history illustrates why the sequence of questions matters: even after a decision on interface protection, the fair-use defence remained to be considered. Jury and appellate proceedings produced different outcomes. Classifying material as protected did not, by itself, complete the assessment of its use. [2]

The legal question and the appropriate test

Section 107 of the US Copyright Act requires consideration of the purpose and character of the use, the nature of the work, the amount and substantiality used, and the effect on the market for or value of the work. It establishes neither a permitted line count nor a safe percentage of copying. These factors structure a contextual assessment; the significance of a portion must be considered in relation to the whole and the circumstances. [3]

Editorial inference: before comparing similarities, it is useful to formulate two separate hypotheses. The first asks whether the material is protected at all. The second asks whether a particular use may be permitted despite that protection. Combining them produces an opinion whose conclusion cannot be reconstructed.

Analytical method: code, function and market

The Supreme Court distinguished declarations used to invoke functions from the implementation that performs those functions. Its decision of 5 April 2021 rested on fair use, assuming copyrightability of the disputed code only for the purposes of argument. It did not finally determine that all APIs are protected or that they fall outside copyright. [1]

Methodological proposal: an audit of a similar project should include a file-provenance table, a map of reused elements, a description of the independently written implementation and an explanation of the technical need for compatibility. Market questions require separate evidence: who uses the solution, for which tasks, and which products it could actually replace. Such documentation makes the reasoning reviewable; it does not guarantee a litigation outcome.

The decision and its reasoning

In these circumstances, the Court’s majority held that the use was fair. Relevant considerations included the interface’s functional nature, the new platform context and the connection between the amount copied and the purpose pursued. The Court also considered market effects and programmers’ ability to apply previously acquired skills. The result did not authorize unrestricted copying of software. [1]

Editorial inference: the case’s most useful lesson concerns defining the subject of the dispute. The general word “code” conceals different technical roles. An analysis should explain which elements provide a means of communicating with software and which constitute a particular implementation of a task. Only then can a compatibility argument be tested against the actual extent of reuse.

Relevance to software projects in Poland

Editorial inference: Google v. Oracle is best read as a study in comparative legal reasoning rather than ready-made instructions for a Polish project. First identify the applicable legal system, type of right, licence terms and user activities. Then compare approaches to documenting function, the need for reuse and economic effects.

A hypothetical software-library audit should produce several separate findings: the origin of the material, how it was used, which contractual conditions were checked and which issues require further assessment. Merely naming a famous precedent does not replace those findings. This is a proposed way of organizing evidence, not an account of work performed by the parties to the historical dispute.

Additional documentation proposal: retain version numbers, checksums and download dates for the files examined. A later reviewer can then reconstruct exactly the same material. Comparing a different release of a library may produce different technical findings even when the product name remains unchanged.

Lessons from the analysis

  • Separate the question of protection from the permissibility of a particular use.
  • Assess the function and provenance of reused elements and their market context.
  • Do not import US fair use automatically into a Polish legal assessment.

Limits of interpretation

This study concerns US copyright and the particular findings in the case. It does not determine the patentability of software or the legality of arbitrary API copying. Methodological inferences are editorial commentary.

Sources and further reading

  1. 1
    Google LLC v. Oracle America, Inc. — US Supreme Court opinion ↗

    18-956; 593 U.S. 1 (2021), 5.04.2021

    Primary source; particularly pp. 11–15 and 21–36 of the majority opinion. The syllabus is a separate editorial summary.

  2. 2
    U.S. Copyright Office — Fair Use Index: Google v. Oracle ↗

    Fair Use Index, Google LLC v. Oracle Am., Inc., 2021

    Official account of the facts and procedural history; separate from the judgment itself.

  3. 3
    Copyright Act — statutory fair-use factors ↗

    17 U.S.C. § 107

    Text of the law published by the U.S. Copyright Office; four assessment factors.

More case studies

All cases →