by David Chang Villacreses, MSc

Disclaimer: David Chang Villacreses is not a lawyer. This document does not include and is not intended to include any legal advice for any purpose whatsoever, and has been written for informational purposes only. When in doubt about best practices on software licensing, please consult your institution’s legal team and intellectual property licensing office.

When Duke innovators come to our office with invention ideas like medical devices, material formulations, or pharmaceutical compounds, both the inventors and our licensing managers are usually on the same page how to proceed: applying for a patent. But for digital innovations, the path forward can fork and loop in different ways – and a patent-based strategy may not always be the best way to go.

When developing software innovations in academic settings, the various components assembled into these projects can be encapsulated under different types of Intellectual Property (IP), depending on each project. The types of IP often used to manage software inventions include: 

Patents are a grant made by a government to confer its owner a limited monopoly to block the use, making, and selling of an invention. In practice, patents are a mechanism to control others’ use of the invention as described under such patent’s claims within a territory. 

For software inventions, patents have become increasingly difficult to achieve. For example, the use of a well-known machine learning algorithm in a new context is generally not enough to be patentable in the US.

IP licensing managers need to be strategic regarding when and how to protect a software invention. For example, if a method is very rapidly evolving, and the claims would become obsolete by the time the patent is issued, or soon after, it may not always be strategic to file a patent for said method.

Another example includes software systems running entirely in the cloud or behind closed doors. If it proves too challenging to detect and confirm infringement of a patent’s claims, patent-based protection strategies may not be worth pursuing.

Copyrights protect tangible outputs of human creativity and are evaluated differently from patent rights. This means that only a modicum of originality needs to exist for copyrights to protect a work of authorship. 

Copyrights are automatically protected, meaning that as soon as authors create or produce a copyrighted work, the copyrights for such works can be licensed in much the same way patents are licensed.

In copyright law, software codebases are viewed as literary work. Similarly, annotated datasets, simulation phantoms, or creative compilations of data are examples of digital innovations that can be protectable under copyrights.

When managing copyrighted works, your licensing office will often want to understand any joint ownership and authorship considerations, as well as strategize if and how derivative works should or should not be allowed.

After submitting a disclosure with a licensing office, said office’s software or digital licensing teams will often look at any third-party dependencies that the authors have used or incorporated. For example, the licensing office may request a list of the open-source libraries, devkits, or other third-party software dependencies that were used in creating the software being disclosed to better understand the licensing strategy for the asset. 

Though each software asset must be supported on a case-by-case basis, and your licensing office can help your team create a licensing strategy for each project, the following best practices and considerations are advisable regarding software inventions:

Please note: each step (development, completion, sharing of code) may have differing strategies, and not all software projects can or should be open-sourced.

It is strongly encouraged to always consult with your institution’s IP licensing office before making any decision to make your codebase openly available. That said, there are various options to open source your code:

When evaluating how you would like to make your software invention available to the public, consulting your institution’s technology transfer office is strongly advised. In most cases, the university owns copyrighted software and materials done in the context of research projects, especially when funded by the government or industry.

Your institution’s technology transfer office can provide a strategic assessment of:

We wish you success in your digital innovation journey, regardless of your institution. But if you are at Duke University, please do reach out to the Digital Innovations team at the Office for Translation & Commercialization.