Data, Data, Everywhere: AI-Supportive Inventions in the Age of ChatGPT

Written By: Liz Gray (IP Lawyer)

With ChatGPT in the news and data privacy and copyright concerns (and lawsuits!) on the rise, it’s worth considering how to protect data collection and generation techniques that support the development of your business’s AI tools!

Did you know that data management and organization tools and synthetic data generation methods can be eligible for patent protection? For example, Brion Raffoul LLP recently drafted and prosecuted US Patent No. 10,726,535. This patent covers systems and methods for generating new training images for an AI system. 

The technique extracts features of interest from a small number of images, determines characteristics of similar features, and recombines those similar features with new backgrounds to create a large number of new images. 

Figure 7 of US 10,726,535 B2, filed in 2019, issued in 2020, drafted and prosecuted by BRION RAFFOUL LLP

Creating training data in this way can help to avoid issues that might result from data collection by more conventional means.  As AI systems become more and more common, the value of protecting AI-supportive tools will only rise!

For more insights on what to patent in AI (and how!), take a look at this paper on Patenting the AI Pipeline by BRION RAFFOUL’s Liz Gray, and contact our team of experts at BRION RAFFOUL LLP.