Home > Our work > Innovation and commercialization > Protecting AI innovation, intellectual property
Protecting AI innovation, intellectual property
Both the development and use of AI technologies have the potential to be hindered by several identified challenges when it comes to intellectual property rights (IPRs). For example: How can we efficiently protect investment through intellectual property protection within a company developing new AI technologies? How can a company’s training data set and pretrained model be protected? What kind of intellectual property rights will be created, and how will ownership of such IP be organized? How can a company address the different jurisdictions of IPRs, and how could this deter innovation?
In 2021, the GPAI experts were able to produce a first version of the GPAI IP Primer v1, which acted as a guide for helping SMEs and startups working with AI to navigate and benefit from intellectual property rights, by addressing:
- Current effective IP laws,
- Best practices, activities, and mechanisms related to IP,
- How individual organizations are currently handling IPs,
- Differences across geographies.
In 2022, the IP Primer v2 was developed to:
- Clarify the differences between trade secret & IP for software industry and how to build a comprehensive strategy; and
- Supplement the IP primer with more guidance on open source and licensing especially in relation with the reuse of trained models.
Reports
2023
2022
Intellectual Property Expert Guidelines for scraping or collecting publicly accessible data (November 2022)
Intellectual Property Expert preliminary report on data and AI model licensing (November 2022)
GPAI Intellectual Property (IP) Primer (November 2022)
2021
GPAI intellectual property (IP) primer (November 2021)