Last week, the UK Supreme Court heard arguments in the case of Thaler (Appellant) v Comptroller-General of Patents, Designs and Trademarks (Respondent). The case concerns DABUS (Device for the Autonomous Bootstrapping of Unified Sentience), an AI system that is alleged to have made two inventions that are the subject of UK patent applications GB1816909.4 and GB1818161.0.
Enlightened arguments from both parties at the UK Supreme Court made for interesting viewing, says partner and AI Report author, Mike Williams in leading IP publication World IP Review today.
But what does it mean for something to be obvious to a machine? This affects the central tenets of the patent system.
It’s also interesting to note what happens in the US, where an AI system cannot be named as an inventor (CAFC 2021-2347). In the US deliberately making a false declaration of inventorship is punishable by imprisonment; would the discrepancy required by a different UK approach stop people filing in the UK? An internationally coordinated approach is required, argued the comptroller.
It will be interesting to read the final decision. It seems likely that the UK Supreme Court will decide, as the US and EPO have, that an inventor must be a human and cannot be an AI system—at least for now.