The world of software innovation is evolving faster than ever, and Tech Transfer professionals are being asked to make critical decisions around open source, copyright, patents, and AI. In this episode, we take a deep dive into this complex landscape with Dan Dardani, Director of Physical Sciences and Digital Innovations Licensing and Corporate Alliances at Duke University. Dan brings over two decades of experience to the conversation, including nearly 20 years at MIT and his long standing leadership in AUTM’s Software Course Committee.

Dan shares his practical insights into how Tech Transfer offices can navigate the tricky decision between copyright and patenting software, how to handle open-source licensing in research environments, and what to consider when commercializing digital tools. He also offers guidance on machine learning and artificial intelligence, addressing real-world questions around IP ownership, data licensing, and the legal gray areas created by emerging technologies.

Whether you’re just starting to build policies for software disclosures and AI-related inventions, or you’re looking to refine your office’s strategy, Dan’s perspective is both grounded and forward-thinking. His advice, drawn from decades of hands-on experience, will help you understand not just what’s changing in the digital IP world, but how to keep up and lead through it.

In This Episode:

[01:57] Patents and copyrights aren't mutually exclusive. They protect different aspects of the software. Software comes with copyright right out of the box.

[02:44] Patents are more difficult and more expensive to obtain. Courts have recently made it more difficult, but it still can be done.

[03:23] When deciding between copyright or patent, ask what the innovation is. Is it a transformative leap? 

[04:57] You need to tell a compelling story and emphasize the transformative ability of your software.

[05:53] The last thing to consider is if there's a commercialization strategy.

[06:48] Examples of software innovations that have met the criteria for patentability. Diamond v. Diehr in the 1980s. Transformative is a key concept in IP thinking.

[08:03] We have to be more careful with applying for algorithms now. 

[09:11] Workhorse apps and code may be more suited for copyright protection than patents.

[10:22] Copyright is the first line measure for protecting innovation.

[10:46] Open-source is vital to software innovation. Risks dealing with open source innovations include third-party code issues, sponsorship issues and open source compliance issues.

[11:45] It's important to not commit copyright infringement by releasing someone else's code.

[13:34] Balancing a researcher's desire to use open source licenses and the universities need to protect IP and pursue commercialization. Education and early communication.

[15:49] There are multiple ways to license, including dual licensing strategies.

[16:14] An example of FFTW using a hybrid licensing model.

[18:13] Releasing code as part of the peer review process.

[21:55] Focusing on machine learning and AI.

[22:07] Addressing IP ownership when working with these technologies. Understanding the difference between being a data producer and a data user.

[23:42] It's crucial to understand the layers and document the data sources.

[24:24] Navigating inventorship when AI is involved. A human needs to be named the inventor.

[26:01] There's going to be an evolution of the laws regarding patents and AI. The laws are going to need to adapt to address inventorship and ownership.

[28:26] Advice for TTOs to mitigate risk that might infringe on existing IP. It gets complicated, and the best advice is to start with a clean house.

[31:13] Distinguishing between the types of data used for licensing.

[33:49] Advice for tech transfer offices that are just beginning to think about these issues and develop policies around proper software hygiene and AI related inventions.

Resources: 

Daniel Dardani - Duke University

Daniel Dardani - LinkedIn

Innovation Without Borders: Insights from the ISTA Forum 2024

Diamond v. Diehr

FFTW