The UK IPO conducted a consultation at the end of last year, following which the Government has announced its response (at a recent UK AI event). Of particular interest is the focus on AI generated inventions, and the question mark over whether new legislation is required to protect AI-generated inventions (where often the usual criteria for patentability are not met).
It is clear from the response that there is little appetite for the AI itself owning its IP (which is quite right for so long as we are not really talking about an actual AI, but rather advanced machine learning). However, there was less consensus when one went beyond that question, with there being a need to do something to ensure that investment in AI is not diminished due to an unclear path to protection.
The question of copyright is also interesting from two perspectives. First, is it attracted to an AI-generated work. Second, and perhaps more importantly, what about the data used to train AI systems? There are question marks over what data may be used for that purpose (though licensing is generally always an option), and consequentially where money is saved may that have an impact, and perhaps create a bias, in the way the AI is trained?
That all said, the key question on my lips is would the DABUS case turn out slightly differently under a future regime?