AI and authors: What comes next

AI relies on the written works of members like you to function. Despite this, technology companies are exploiting authors’ works without permission, recognition or compensation and face no consequences for doing so. We understand that many of you are rightly concerned and want us to explore possible solutions.

We’re launching a survey to gather your views on both AI generally and on two potential licensing solutions. We exist to represent you and we need to make sure we get this right, so I urge all members to participate. Any solutions must have the support of our members and your collective views will influence the options we pursue.

Technology companies like OpenAI are currently attracting huge investments for products like ChatGPT, which in part rely on the use of written works without the consent of their creators. While there are options being developed to allow rightsholders to opt their works out of being used to train AI systems, currently there are no proposals from the large tech companies to address permission from or compensation to creators.

Copyright Licensing Agency (CLA), the company who licence organisations to copy your published works, recently launched their Text and Data Mining (TDM) licence, which allows companies to extract insights and patterns from published materials. Although generative AI systems do rely on TDM techniques, this licence strictly prohibits this type of use. Instead, the CLA is currently developing two separate collective licensing solutions to remunerate creators for the use of their works within generative AI systems.

One of these would allow workplaces to purchase licences, allowing them to use extracts from works in the CLA repertoire as prompts for generative AI systems to produce outputs, such as content summaries. This licence is an evolution of the use of your work from photocopying and scanning into the technology of today, as more workforces begin to incorporate tools like Microsoft Copilot into their routine work patterns. The other licence would be offered to technology companies, allowing them to use published works to ‘train’ their AI models. We have explained these proposed licenses in more detail here.

Some of you may have already been contacted by your publishers to ask you to sign addenda to your contracts which would potentially give you some form of compensation for the use of your work to train AI systems, akin to the CLA’s proposed training licence. What you choose to do with your works is entirely your own choice, however we strongly believe that the best option for writers is a collective solution.

What you choose to do with your works is entirely your own choice, however we strongly believe that the best option for writers is a collective solution.

There are a number of reasons for this, which we summarised in our recent statement on AI. Firstly, we know many of you will have produced works with different publishers. This would mean negotiating and administering separate deals for each publisher. A licensing deal through one collecting society would be far simpler for writers to administer.

Secondly, we believe that collective bargaining would give us a strong negotiating position and allow us to achieve better terms for writers. Finally, a collective solution ensures that writers, visual artists and publishers would all be remunerated fairly. All too often, creators are not given their fair share, and a collective solution would allow for all creators to be included.

We know that concerns around generative AI and its impact on creators are widespread. Meta (formerly Facebook) has caused alarm with the unveiling of their new privacy policy, which means users of their products agree for their data to be used in the development of AI models. Our colleagues at DACS explain this in great detail and offer advice about what you can do about it. Similarly, the Society of Authors has published practical advice about how you can protect yourself and your works from this challenge.

I recently participated in the CISAC (an international non-profit for creator rights) General Assembly in Seoul, and it was heartening to see how seriously these issues are being taken around the world and fascinating to see the different policy solutions being developed.

Here in the UK, we are of course eagerly anticipating the General Election on 4 July. The UK has yet to set out a coherent position on AI regulation, however, we expect this to be high on the political agenda for a new government. We have heard some positive noises from both Government and opposition figures and will be doing all we can to ensure that future policies serve the interests of creators and respect their rights. You can read more about the current policy and legislative landscape here.

For 47 years, we have ensured that writers are recognised and remunerated wherever their works are used. During this time, we have adapted to profound technological changes. In the 1970s it was photocopiers, in the early 2000s it was scanning and online use and now we are seeing a new shift with the development of generative AI. Amidst all this change, one thing is certain: where your rights are concerned, ALCS will continue to strive for recognition, respect and remuneration.

Barbara Hayes
Chief Executive, ALCS