I’m doing things a bit backwards this week, I’m writing this article about the apparent lack of software (and specifically AI-related) IP to license and the reasons behind that first, and then, I’m posting a PoW that’s an AI-related one on Tuesday. Despite the fact that I’m sharing this patent, I’m personally of the opinion that licensing software IP is notoriously challenging and might only make sense in a very limited set of circumstances.
A bit of history
Let’s start with a bit of history, the history of software traces its roots back to the early days of computing in the mid-20th century. The concept of software emerged with the creation of the first programmable computers in the 1940s. Early pioneers like Ada Lovelace envisioned machines that could be instructed to perform various tasks through coded instructions. During the 1950s and 1960s, programming languages such as Fortran and COBOL were developed, laying the foundation for software development. Up to this point, the concept of patenting software didn’t exist at all. Software was perceived as a technical method on top of hardware solutions that were being patented.
The advent of personal computers in the 1970s and 1980s, spearheaded by companies like Apple and Microsoft, revolutionized software, making it accessible to the masses. This era introduced the concept of patentable software as the legal landscape expanded to cover software as a separate category. As a backlash to software patents in that era, the open source movement was born and it pretty quickly demonstrated the power of collaborative independent developers and teams working on improvement software together. The open source movement has been a big reason why trivial and wide software patents stopped being effectual; this was a really important step in the evolution of software patenting.
Why it doesn’t make sense to patent software now
There are a handful of reasons why this is the case so let me focus on the top 3:
Short Innovation Cycles. Unlike hardware, software is an ever evolving technology and what you patent today, might change tomorrow. Considering the process of filing patents can take months to years, it doesn’t make economic sense to hold on to a specific version of your software and patent that.
Trade secrets are more effective. Just like KFC and Coke don’t disclose their secret recipe, you’re often better off as a software inventor to build a business with. your solution rather than disclosing it where anyone can reverse engineering and enforcing your patent becomes a very expensive proposition.
Risk of patent trolls. The software industry is particularly vulnerable to "patent trolls" – entities that acquire patents solely to extract fees from companies through litigation or threats of litigation. This risk can discourage companies from pursuing patents to avoid drawing attention to their innovations.
When does it make sense to explore software IP?
Software can be a broad term, so let’s focus specifically on algorithms. If you have discovered a novel algorithm from a lab with deep expertise in specialized fields like weather modeling, financial forecasting, or medical imaging, licensing could be worth considering. These areas are exceptionally specialized, and having exclusive access to deep domain knowledge can provide a significant competitive edge.
With that being said, it is crucial to make sure you have a strong, ongoing relationship with the lab. This will guarantee access to future innovations related to your patent, allowing you to continually evolve and stay ahead without the risk of your technology becoming obsolete.
About the artist
Omyo Cho, a South Korean artist born in 1984, is gaining recognition for her unique blend of science fiction storytelling and sculptural works.