AI and IP – the UK government's response to the UKIPO's consultation

Quick read

Here's an overview of the "soft" IP aspects of the UK government's response to the UKIPO's consultation on AI and IP, and why IP rightsholders should care about these developments.

How did we get here?

Last year, the UKIPO launched a call for views to understand the AI implications for IP and vice-versa (see the questions summary here). 92 stakeholders replied, including from the tech sector, creative industries, trade bodies and industry associations (see the responses here). If you weren't one of them, we hope you might change your mind after reading this.

What's the development?

The UK government recently set out its action plan in response to the 92 replies.

Copyright and related rights

The government will review how copyright owners license their works for AI use and how this can be facilitated, including by improving licensing or through copyright exceptions aimed to support innovation and research.

The government will also consult on whether it should replace the copyright protection for computer-generated works with a related right that would reflect the investment in such works. It will consider what action it could take to reduce the risk of false-attribution (human versus AI works) and whether only original human creations (including AI-assisted ones) should benefit from copyright protection.

Trade marks

One of the issues up for consultation was the potential impact of AI on the "average consumer" concept – because AI systems can make purchasing suggestions and reduce brand interaction with consumers – and trade mark law more generally (for example, the possible increased relevance of post-sale versus point-of-sale confusion and (where voice-assisted tech is used) of the aural comparison). The government also consulted on the impact of AI on infringement (do AI actions amount to trade mark infringement?) and liability (if yes, who should be liable?).

The government contends that the current trade mark legislation is fit for purpose (there is not enough to impact the core legal concepts yet). While AI systems may infringe trade marks, there is little to no impact on the trade mark infringement provisions and the courts can explore those issues on a case-by-case basis. So, while no action is in the pipeline now, the government intends to monitor the issues as AI tech develops and remains receptive to further views on trade marks and AI.

Designs

As for trade marks, the government considers the current legislation generally fit for purpose (in particular, the "informed user" test applies to AI-generated designs) but will keep the position under review. The government agrees with stakeholders that, while AI systems should not be authors or design owners, they may be used to infringe designs. Still, longer-term questions remain (and engagement with interested stakeholders will continue):

Who is the author of an AI-generated design? (The government notes the similar approach to copyright law regarding computer-generated works and it will ensure that the review of the copyright provisions will take this into account.)
Should an individual be deemed liable for AI infringing acts?
Can the established legal tests remain relevant for AI-generated designs?

Trade secrets

There's nothing on the horizon here, but trade secrets protection will be further explored alongside patent developments.

Why should IP rightsholders care?

There is a real opportunity here for public policy engagement that will impact what form UK IP rights might take where AI is involved, what licensing mechanism should be used for AI development and what role, if any, should AI have in IP enforcement (the UKIPO commissioned researched into this, which is expected this autumn).

While the UKIPO consultation is now closed, the UKIPO will still collaborate with business and tech experts to develop a "robust evidence base" and will communicate and engage with IP owners. The UKIPO will also continue to see how it can integrate AI operationally as part of its transformation programme (see the One IPO Transformation Prospectus, which builds upon the trade marks Pre-Apply service) and will use AI for data validation and making it available for business use.

Is there a bigger picture?

Yes. The UK government is not alone in grappling with AI, although it seems to have lagged behind the EU. On 21 April, the European Commission proposed a regulation on a European approach for Artificial Intelligence which is only one piece of the EU's vision, set out in the European Commission's AI White Paper, the European Strategy for Data and the EU IP Action Plan.

It is unsurprising that the House of Lords Liaison Committee reported that "[t]here must be more and better coordination, and it must start at the top". The UK government is currently considering the AI Council's Roadmap recommendation for the UK to adopt a national written strategy on AI.