On 4 December 2020, I have talked about AI and patent law at the 2-day workshop Neuroni Artificiali e Biologici organised by the University of Trento. This is part of the ERC-funded project BACKUP led Professor Lorenzo Pavesi, Chair of Experimental Physics. BACKUP will address the fundamental question of which is the role of neuron activity and plasticity in information elaboration and storage in the brain.
This workshop assessed the legal, philosophical, and ethical issus in AI, robotics, and biotechnologies. Its goal was to start a multidisciplinary conversation to set up a community where we can identify paths for future collaborative research.
The workshop consisted of talks, including mine, roundtable conversations, and a white paper drafted by the PhD students in attendance.
Key takeaways of my talk
We must foster an inclusive and diverse public debate on AI & ethics. This has also practical consequences because immoral AI applications cannot be patented
The widespread use of AI applications to carry out research and produce inventions leads to the risk of over-monopolisation of ideas thus stifling innovation. A way to address the issue is to change the standard to assess the investive step, that is a requirement to meet for an invention to be patentable. Inventive step is assessed from the perspective of the person skilled in the art. This low threshold should be replaced with the higher threshold of the ‘AI-enhanced multidisciplinary team’.
The European Patent Office, the UK Intellectual Property Office and the US Patent and Trademark Office agree that AI cannot be an inventor. Fair enough, but how do we prevent human inventors to make-up humanity? How do we verify that an invention that is presented as human-made is not actually AI-generated? Could an AI system verify humanity?