In November 2021, we reported on a consultation by the UK Intellectual Property Office (“UKIPO”) on potential amendments to existing intellectual property law in light of AI created works and inventions. The consultation sought responses on three key issues:
- copyright protection for computer-generated works without a human author;
- licensing or exceptions to copyright for text and data mining; and
- patent protection for AI-devised inventions.
The UKIPO has since reviewed all feedback received and, on 28 June, published their consultation response in the latest effort to shape the framework of intellectual property and AI within the UK.
The Consultation Response
The consultation response continues to highlight the ability of AI to support innovation and creativity in a number of forms, and its necessity as a tool to be used by the UK Government to secure itself as a global technology superpower. Without a suitable regime to protect their IP rights, currently valued at an estimated £63 billion within the UK, rights holders would be less incentivised to continue to innovate and create. A balance must therefore be struck, and this was a recurring theme throughout the response.
Issue 1: Copyright in computer-generated works:
As part of the consultation, stakeholders were asked whether computer-generated works (“CGW”) (those created without a human author) should continue to be protected under UK copyright law, and if so, how. CGWs are currently protected for 50 years.
Three options were given in order to direct opinions:
- no change to the law;
- remove protection for CGWs; or
- replace the current protection with a new reduced right with a narrower scope or duration.
Most respondents indicated that the current law was sufficient and no changes were needed at the moment.
Action: As a result, the UKIPO has elected to make no changes to the law, citing the lack of evidence that the current mechanisms in place are harmful. It did, however, note that as AI advances, further evaluation may be required, and amendments made, if the evidence supports it.
Issue 2: Text and Data Mining
Text and data mining (“TDM”) (the process of using computational techniques to analyse large quantities of data) which is used, among other things, when training AI systems was also considered. While factual data, trends, and concepts, are not protected by copyright, they are often embedded in copyright-protected works. Unless permitted by virtue of a licence, or by one of the very limited exceptions, use of such data in data mining would result in copyright infringement.
The consultation sought views on how to make it easier for people to data mine copyright materials, while still protecting the rights of the holder.
Stakeholders were asked to consider three broad options:
- make no change to the current regime (which essentially only permits TDM for non-commercial research purposes);
- improve licensing for the purposes of TDM; or
- implement a number of potential exemptions for the use of TDM.
The majority of rights holders selected no change or licensing solutions to make more material available for TDM. Users, unsurprisingly, favoured a wider exception as the preferred option, citing the cost of licensing and difficulty in obtaining licences.
Action: The UKIPO has elected to introduce a new copyright and database exception, permitting TDM for any purpose, without an ability to opt (or contract) out. This was considered to be the most supportive of AI and wider innovation and consistent with the UK Government’s ambition to make the UK a global centre for AI innovation. It was, however, restated for clarity that rights holders will still have safeguards protecting their content such as the requirement for lawful access. Rights holders will be able to choose the platform where they make their works available and charge for access.
Issue 3: Patents and AI
The final issue addressed by the response pertains to patent law and AI. A number of recent developments, such as the efforts of Dr Stephen Thaler to have an AI system recognised as a patent inventor, have led many to consider whether the current patent regime is still appropriate.
As with the other issues, Respondents were offered the options of:
- no legal changes to the current regime;
- expanding the definition of inventor;
- allowing AI to be named as inventor or remove requirements to name an AI inventor; or
- creating a new patent-like right to protect inventions devised by AI.
Many respondents were concerned that any change to the law should be at international, rather than national, level, to avoid prejudicing UK interests. Any change to the law on inventorship needs to be harmonised internationally. The consensus was therefore to maintain UK law as it is (for now).
The response, highlighted four recurring reasons for this consensus:
- there is still no consensus on who should own patents when AI is the inventor;
- the technological case for change may be stronger in the future;
- international harmonisation would be required as a prerequisite to successful change; and
- a unilateral change to UK legislation would be premature and may be negative for the UK’s innovation landscape.
Action: For now, the UKIPO has elected to make no change to current UK patent law to limit the risk of creating additional barriers to innovation and investment in AI technology. It was, however, noted that this item would remain open for further discussion and may be revisited in future. In particular, the UKIPO should remain at the forefront of international harmonisation discussions on AI inventorship and address any (incorrect) perceptions that the current law prevents the patenting of AI-assisted inventions.
DLA Piper continues to monitor the UKIPO’s updates and developments in relation to AI and intellectual Property. For further information or if you have any questions please contact the author or your usual DLA Piper contact.