Following the UKIPO’s 2021 consultation on how artificial intelligence should be addressed within the patent and copyright systems, the UK Government has published its response. While determining not to pursue any changes regarding patent inventorship or copyright protection for copyright generated-works at this stage, the UK Government will move to amend copyright law to facilitate the easier analysis of material for machine learning, research and innovation.
As discussed in our blog post in December 2021, the UK Intellectual Property Office’s (the UKIPO) consultation (the Consultation) was conducted as part of the Government’s plan to be the “the most pro-tech Government ever” and focused on three key areas:
- licensing or exceptions to copyright for text and data mining (TDM), which is often significant in AI use and development;
- copyright protection for computer-generated works (CGWs) without a human author; and
- patent protection for AI-devised inventions.
Following the Consultation, the Government has determined that it should change the law in relation to item (1) TDM, but not for items (2) or (3) regarding protection for non-human authors or inventors.
Licensing or exceptions to copyright for text and data mining
As explained in our December 2021 blog, TDM means “the use of automated computational techniques to analyse large amounts of information to identify patterns, trends and other useful information” and is used in training AI, marketing, and business analytics amongst other things. TDM systems copy works to extract and analyse the underlying data and, under current law, this copying constitutes copyright infringement unless permitted under licence or an exception (such as the non-commercial research exception, introduced in 2014).
With a view to making it easier to carry out TDM and therefore support AI and innovation, five policy options under consideration in the Consultation:
- Maintenance of the status quo (ie an exception to copyright for TDM in the case of non-commercial research) if it is shown that the current law is not hindering access to material for text and data mining, particularly in the context of training AI systems.
- Improving the licensing environment by creating model licences and codes of practice for copyright licensing in the context of text and data mining.
- Extending the copyright exception to cover commercial research and databases.
- A new exception for text and data mining conducted for any purpose by anyone, with an opt-out system for rights holders to carve out their works. Such an exception would not be restricted to scientific research and could include, for example, journalism or business analytics. Although, researchers would still need lawful access to read the copyrighted works and databases in the first instance. This approach, it is suggested, would remove the high costs associated with mining works for which it is difficult to locate the relevant rights holders while allowing rights holders to license their works if they wish.
- A new exception for text and data mining conducted for any purpose by anyone without an opt-out system. Although, as above, researchers would still need lawful access to read the copyrighted works and databases, rights holders would not be able to prevent text and data mining of these works and databases.
A range of responses were received, with some considering there to be no issue with the current system; others noting that TDM for any commercial purposes justified rights holders being able to charge for a licence; others put forward different licensing models to aid access; whilst others considered that introducing exceptions would reduce the costs of TDM and avoid pricing users out of the market. As one would expect, rights holders were broadly against the proposal to expand the scope of the exception.
The Government concludes that it will change the law governing TDM, to introduce a copyright and database right exception allowing TDM for any purpose (including commercial exploitation) such that rights holders will no longer be able to charge for UK licences for TDM.
The Government emphasised that rights holders would still have safeguards to protect their content, including requiring users of TDM to have lawful access to the content. Rights holders will be able to choose the platform where they make their works available and can charge for access via a subscription or single charge. The Government also noted that certain countries in the EU, as well as Japan, Singapore and the US (under fair use law) already had copyright exceptions for TDM to varying degrees.Whilst some respondents to the Consultation raised data protection concerns including regarding data ethics and standards these considerations fell outside the scope of the Consultation.
Copyright protection for computer-generated works
The Government will not change the law as it concerns CGWs, an approach which is in line with the majority response in the Consultation.
Unusually, CGWs in the UK are currently protected under copyright law but the UKIPO found no evidence that this current protection is harmful. Moreover, the use of AI in CGWs is still in its infancy, meaning that the Government did not think that it could properly evaluate the impact of removing or providing for alternative protection - given this lack of information, the Government was concerned about the risk of unintended consequences. Further, most respondents thought that protection for CGWs was not currently widely used in any event. The Government will further monitor the impact of existing protections on AI.
Patent protection for AI-devised inventions
The Government will not change the law as it concerns patents and AI inventorship, again, mirroring the view of the majority of respondents to the Consultation.
The Government found no evidence that current UK patent law (requiring a human to be identified as an inventor) is unsuitable to protect inventions made using AI, particularly as many respondents felt that AI was not yet sufficiently developed to invent without human intervention. The majority of respondents considered that there was limited technological, legal, and economic evidence for legal change. The Government was also sensitive to many respondents’ concerns that future changes should be done at the international level to ensure harmonisation, pertinent as we see cases across the globe considering the inventorship rights of AI and whether, for example, DABUS or Dr Thaler should be identified as the inventor (our blogs here and here consider this topic further).
Other issues raised by respondents
Interestingly, many respondents raised their concern that there needs to be a broader debate as to whether AI should be recognised as a legal personality but this issue was outside the scope of the Consultation.Some respondents also voiced their concerns regarding the potential negative impacts of AI on performers. For example, there is the concern that computer-generated performances may either replace human performers or may be used to use their likeness without their permission ('deepfakes'). The Government stated that it would keep these issues under review.
The pace of AI advancements continues to drive legislative developments. Notably, the Government emphasised in its response that it would continue to review and consider further changes in several areas of the Consultation where it proposes not to take action for the time being.Another theme in the Government's response is the UK's willingness to develop its intellectual property framework following Brexit, with a view further to facilitate innovation. As in related areas, such as data protection, it remains to be seen how far the UK will diverge from the EU. However, it is clear that we should expect further developments in this field.