AI as an Inventing tool—Its Implications for Patent Law and Policy

Sponsor: BCLT

With the remarkable progress of deep-learning-based AI, we are witnessing a paradigm shift in the inventing process itself. A good example is AlphaFold’s accurate prediction of 3D structure of proteins, and the subsequent application of its descendants to drug discovery. For the first time in history of human inventions, a tool is capable of predicting a definite and permanent solution, if the problem space has been properly defined. While this new synergy between human researchers and machine (AI) is likely to unleash unprecedented potential in the innovative economy, it may pose challenges to various patent law doctrines, including inventorship, novelty, non-obviousness, enablement and written description. In dealing with these doctrinal challenges, a holistic approach undergirded by deeper principles is in dire need. Meanwhile, AI can also be used as a useful tool to promote fair access to the patent system. This half-day conference invites several AI technologists and pioneering legal scholars to share their insights on these important issues.

---