It is a busy time at the moment for IP offices and other institutions, as they consider and consult on the implications of artificial intelligence (AI) on IP. In March 2021, the UK published the results of its call for views, and from this, it has developed a list of priorities that it will take forward.
Not surprisingly, the issues of protection for inventions and original works were key for the respondents, together with the closely related topic of ownership. Although there was consensus that AI itself should not own IP rights, there were different opinions on whether inventions or works created by AI should be protected.
As a consequence, the government has said that it will consult later this year on policy and legislative change for protecting AI generated inventions which would not otherwise meet the inventorship criteria. It will also consult on whether to limit copyright to human creations, including AI-assisted creations. This would entail a consideration of whether to replace the existing protection for computer-generated works (a highly unusual provision) with a related right to reward the investment made in such works.
Another big issue for respondents was training data. Many recognised the importance of copyright-protected material in training AI systems, but there were diverging views on whether there was adequate access to such material. Another point for consultation will therefore be on measures to make access easier, which will include reviewing the text and data mining (TDM) exception. This is a contentious issue as demonstrated in the negotiations over the TDM exceptions in the Digital Single Market Directive.
However, because of Brexit, the UK government will not be implementing this directive, which means that at present only a very narrow TDM exception applies in the UK. The government has also said that, in the patent field, it will consider whether a deposit system for training data disclosed within patent applications would be feasible and beneficial.
After inventorship, the main concerns expressed in answer to the patent questions were on patent exclusion. Many pointed out that the exclusions make it difficult to protect developments in AI systems themselves and that the more permissive approach of the EPO gave a better outcome. Consequently, the IPO will publish enhanced guidelines on patent exclusions and will conduct a review of any differences in outcome for AI patent applications as between the EPO and IPO.
In contrast to patents and copyright, respondents were generally of the view that the law in relation to trademarks, trade secrets and designs is adequate and flexible enough to respond to existing challenges of AI.
Many of the issues raised in the call for views were familiar from other consultations. Since 2019, WIPO has been conducting a wide-ranging ‘conversation’ on AI and IP which encompasses, not only IP policy but also AI's use in the administration of the IP system. Last October, the USPTO reported that, in very general terms, the majority of respondents to its requests for comments expressed a sense that the existing US IP laws are correctly calibrated to address the evolution of AI. Similarly, a report prepared for the European Commission published last September concluded that there was no immediate need for substantive changes in copyright and patent law to meet the current challenges of AI, although further study may follow a subsequent report from the European Parliament.
In this current climate, it will be fascinating to see whether the UK will propose any policy or legislative changes.
The material on this site is for law firms, companies and other IP specialists. It is for information only. Please read our Terms and Conditions and Privacy Notice before using the site. All material subject to strictly enforced copyright laws.
© 2021 Euromoney Institutional Investor PLC. For help please see our FAQs.