Following a recent Intellectual Property Office (IPO) consultation, it has been decided that AI will not be allowed to patent inventions.
As highlighted by a legal battle dating back to 2018, some people believe that if AI systems invent something, they should be granted the status of being listed as the inventor in the patent for the invention. So for example, Stephen Thaler, creator of an AI system called Dabus (which invented a food container with enhanced insulation & stacking properties and an attention-grabbing flashing light for emergencies), listed Dabus as the creator of the inventions when patents were filed for them back in 2018. This led to a court case (and an appeal) where a UK panel ruled by a two-to-one that majority that an inventor must be a real human person under UK law, not an AI system.
The recent IPO Consultation document (online) agrees that for AI-devised inventions, “we plan no change to UK patent law now. This is because most respondents felt that AI is not yet advanced enough to invent without human intervention.” Also, the IPO Consultation notes that, of the world’s countries, the “overwhelming majority, including the UK, restrict patent inventorship to natural persons” and that “UK patent applicants must name a human inventor or inventors. This was confirmed by the Court of Appeal.”
Although it has been decided that at the current time, in the UK, AI systems can’t be named in patents as the inventor, this view may change. For example, the IPO’s consultation document says that “we will keep this area of law under review to ensure that the UK patent system supports AI innovation and the use of AI in the UK” and that “we will seek to advance AI inventorship discussions internationally to support UK economic interests”.
Computer-generated works (CGWs) are copyright works without a human author and are currently protected in UK copyright law. The IPO’s consultation document states that there is no evidence that protection for CGWs is harmful and that it plans no changes to the law for CGWs, although this law will be reviewed.
TDM refers to using computational techniques to analyse large amounts of information to identify patterns, trends and other helpful information. TDM is used to train AI systems in journalism, marketing, business analytics and cultural heritage organisations. One of the main copyright issues with TDM is that, although factual data, trends, and concepts are not protected by copyright, they are often embedded in-copyright works. Also, some rights holders license their jobs to allow TDM, but others do not, and this can mean financial costs for people using data mining software.
The UK government is keen to make it easier for people to data mine copyright materials in order to support AI and wider innovation in the UK. Also, the IPO’s consultation revealed that the costs of licensing are high, and there are difficulties in obtaining licences (especially when many rights holders are involved). Therefore, the IPO has confirmed that it will introduce a new copyright and database exception, which allows TDM for any purpose.
It is interesting also to note that, back in April 2021, the performing arts workers union ‘Equity’ launched a “Stop AI Stealing the Show” campaign, which called for changes in the copyright law to stop AI from using samples of an actors’ voices or faces to generate content such a deep fakes. Equity was (and still is) worried that current copyright laws make it too easy for AI systems to replace skilled professional performers in many different areas, including automated audiobooks and digital avatars.
The fact that there is the need for an IPO Consultation of this kind about AI shows that it is making incursions into many areas. As it develops and improves, this will continue, hence the need to keep some of the copyright laws under review. Many skilled people in business will be aware of AI tools that can now be used to perform tasks that were part of their jobs before, and it is not surprising that Equity has been seeking more copyright protection against AI. The government clearly want to be seen as “the most pro-tech Government ever” and, as such, appears to be looking to change things in favour of AI where possible. However, there is an argument that AI and its applications are advancing too far ahead of regulation. For the moment, most countries, including the UK, still think that an AI system cannot, in reality, be the sole inventor of something as named on patents because human involvement is needed. Still, the rate at which AI is advancing and the fact that this law remains under review (in the UK) suggests that this view could change in the near future.
