UK Government abandons plans for voluntary code of conduct for AI

It is terrifying to think, considering how pervasive it now seems to be, that artificial intelligence (AI) exploded into the public consciousness only a couple of years ago. Less surprising are the copyright concerns and questions that swiftly followed. The recent UK Government response to the White Paper consultation confirmed that a voluntary code is not going to happen, and instead, an approach of delay and indecision seems to be the current default setting.

This is a pattern we have encountered before. The emergence of new technology over the last 20 years has been accompanied by debate as to how to regulate its emergence and deal with the issues, often IP, typically copyright, related. Mostly, such occasions have been accompanied by hesitant regulatory catch-ups and the resultant patchwork of regulations and amendments to the statute. There is, typically, a reluctance to step in and regulate lest it stifle innovation and leave the UK behind. The issue is that this regulatory hesitancy allows the emerging tech to gain a position of strength. It becomes “indispensable” and weakens the resultant regulatory discussion. This comes through, for example, threats to cut off services if regulation is pursued, at pace, or otherwise.

Turning back to AI and IP regulation, court cases are already well publicized and underway, especially in the UK and the US, asking whether the process of training AI machines through data trawls engages copyright infringement law. Those cases are closely watched, but are they already too late? The AI cat is out of the bag and stuffing it back in is not necessarily an option.

With the very different interests involved, as well as opposing positions, it is not surprising that attempts to create a voluntary code have collapsed. The positions of the parties are too far apart at present and, while both arguably have an interest in agreeing to regulation to ensure a smooth path ahead, the need for an agreement does not seem to have crystalized for the AI parties, even if it is crucial for the rights holders. While the AI parties can look back and see the history of tech regulation, why should they be any different? 

History is firmly in favor of permissive technological innovation and this should concern creators pushing for action. 

The difficulty is that even with a positive outcome in the court cases, with every delay AI gains strength, and the regulatory position becomes arguably harder. With these continued delays there is a risk that concerns will ultimately be ignored. How likely is it that these AI programs will be shut out of the UK in the wake of a negative decision? The greater risk is that such decisions result in regulatory intervention in favor of AI to preserve its presence given the desire to encourage innovation and have the UK as an attractive place to introduce new tech.

Given the issues at stake, it would be helpful to see a code revived with regulatory force behind it. Looking at just the IP issues, a comprehensive solution for compensating creators with respect to the training data used in AI would be a bare minimum. 

This could be coupled with clear routes to opt-in or opt-out (more likely) of being used in that data training along with clear rules as to the transparency of the data sets. More thorny is the question of the output. There is a sliding scale of complexity ranging from identical replication in the output all the way through to questions as to what happens with “in the style of” type output.

There is clearly a need for regulation. The courts are not the appropriate forum for what are, ultimately, going to have to be policy-driven decisions. The current approach of more consultation and delay, with a seeming “wait and see” approach, favors only AI. It would be good to see a more proactive approach taken to support the creative industries that feel their creativity is being exploited in the name of progress.

The precedent does not inspire confidence and mirrors the current approach of delay. However, the challenges this time are faster moving. Whilst we’ve been here before, we can only hope that this time will be different!

← Back to News