In April we looked at IP and Artificial Intelligence, based on my recently published IAM article, specifically where ownership sat by country and company. The key takeaways were three fold:
- There are key players leading the AI patent charge (Microsoft, IBM, Facebook, Qualcomm, etc.)
- US and Chinese ventures are leading in terms of AI investments, and will have a huge advantage over other countries R&D direction (Canada, France, UK) in the future.
- China is poised to be well ahead of the US by 2020, and the push is coming from the research institutions
As an AI researcher, startup-venture, or institutional policy maker, how do we address this?
AI researchers: it is critical to ensure a plan of IP protection is discussed as a critical component of both the research plan and commercialization plan. Foresight is needed to ensure relevant patents are filed, and rights are either kept by the research institution for future license opportunities, or freedom to operate for commercial spin-offs. While technology and IP transfer to an industry partner is one option, patent protection with a non-exclusive license may keep future opportunities open to the institution to build on the research in their own commercial setting.
Startup Ventures: As with researchers, foresight through the creation and execution of a patent strategy is critical. This will reduce the freedom to operate risk as the venture scales, as well as ensure intangible asset value is built into the venture for future needs of cross-licensing or M&A.
Policy makers: To ensure some level of returns can be generated for AI based research investments, policy makers must ensure funds allocated to AI based investments are not left as unprotected research. Further, as some level of protection is sought, ownership must be structured to ensure future access to the IP. This could be done through retaining either freedom to operate access, or preferred licence access. This future access is critical to ensure the policy investment can be built on with future investments. Alternately, following the model by Israel’s Office of the Chief Scientist, IP can be acquired and but require monetary level of reimbursement to the policy maker to ensure funds for re-investemetns continue to grow if the technology or R&D roles are transferred out of the Israeli jurisdiction.
In summary, any research group, scale-up, or multinational working and patenting (or not patenting) in AI should be aware of their competitive IP position against the patent pace of both US based corporations and Chinese based research institutions. The future of AI is on pace to be owned by a select few – yet a well crafted and cohesive strategy for a researcher, startup, or policy group has the ability to level the playing field.