In May, Informa, a multi-billion dollar multinational corporation based in the UK, announced in a trading update that it had signed an agreement with Microsoft that includes “access to advanced learning content and data, and a partnership to explore AI expert applications.” Informa is the parent company of Taylor & Francis, which publishes a wide range of academic and technical books and magazines, so the data in question may include the content of these books and magazines.
According to reports published last week, the content authors were apparently never asked about or informed about the contracts, and they claim they were never given the opportunity to opt out of them or receive any money from them.
Academia is just the latest of several groups of so-called content creators who resent having their work included in generative AI models now vying to absorb the products of human culture. Newspapers, visual artists and record labels are already suing AI companies.
It is unclear how Informa will respond to the rumors of dissatisfaction, but the agreement serves as a reminder to authors to be careful about the terms and conditions of any publishing contracts they sign.
What is the deal with Informa?
Informa's latest update lists four focus areas for the Microsoft deal:
- Improving Informa's own productivity
- Development of automatic citation tools
- Developing AI-based research support software (such as the system being tested by the online academic library JSTOR)
- It will provide Microsoft with access to the data to “help improve the relevance and performance of AI systems.”
Informa will be paid more than £8 million (A$15.5 million) for initial access to the data, followed by unspecified ongoing payments over the next three years.
It's unclear exactly what Microsoft plans to do with the data access, but presumably the content of academic books and papers will be added to the training data of generative AI models like ChatGPT. In principle, this should make the output of AI systems more accurate, but existing AI models have faced heavy criticism for not only repeating training data without citations (which could be considered a form of plagiarism), but also for fabricating false information and attributing it to real sources.
However, the update also states that “the agreement protects intellectual property rights, including by limiting verbatim text excerpts and adjusting for the importance of detailed citations.”
The “verbatim text excerpt limitation” mentioned appears to relate to the US fair use doctrine, which permits certain uses of copyrighted material.
Many generative AI companies are currently facing copyright infringement lawsuits over their use of training data, and their defense is likely to rely on fair use claims.
The importance of detailed citations can be linked to the concept of attribution in copyright, which is a moral right that authors have. It stipulates that when a work is reproduced, the creator of the work should be known and attributed as the author.
How does academic publishing typically work?
Most academics do not receive compensation or benefit from publishing scholarly articles. Rather, the writing of journal and conference papers is usually considered part of the work that falls within the scope of a full-time, tenured position. Publication enhances a scholar's credibility and promotes their research.
The basic process often goes like this: an author researches and writes an original paper, then submits it to a journal publisher for peer review. Most reviewers and editorial board members also do not receive compensation for their work.
In fact, some journals may require authors to pay an “article processing fee” to cover editing and other costs, which can run into the thousands of dollars for open access publications. Generally, the more prestigious the publication, the higher the fee.
If the article passes peer review, authors will be asked to sign a publishing agreement, the terms of which may include logistical arrangements such as when the article will be published, the format (print, online, or both), and royalty distribution (if applicable), as well as arrangements regarding copyright and ownership of the article.
Authors are typically required to grant the publisher exclusive rights to distribute and publish their paper, meaning they cannot publish it elsewhere, and the publisher may be able to sublicense the paper to third parties, such as AI companies.
Some publishers may require authors to transfer copyright to their articles through a permanent copyright transfer agreement.
It essentially means that the author transfers all of their rights as copyright holders in the work to the publisher, who can then freely copy, communicate, distribute, or license the work to others.
It is possible to assign only limited rights rather than all rights, and this is something authors should take into consideration.
Content Mining
It is important that authors understand the implications of licenses and assignments and think about exactly what they are agreeing to when signing contracts. Given the recent trend of publishers entering into agreements with generative AI companies, publishers' AI policies should also be scrutinized closely.
A standard collective licensing solution for the use of content in in-house AI systems was recently released in the US, outlining rights and remuneration for copyright holders, and a similar license for the use of content in AI systems is likely to be introduced to the Australian market soon.
The types of contracts being signed between academic publishers and AI companies raise big-picture concerns for many academics: Do we want our academic research to be reduced to content for AI knowledge mining? There are no clear answers about the ethics and morality of such practices.
Wherett-Potter does not work for, consult, own shares in, or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond his academic appointment.
/Courtesy of The Conversation. This material by the originating organization/author may be out of date and has been edited for clarity, style and length. Mirage.News does not take any organizational stance or position, and all views, positions and conclusions expressed here are solely those of the authors.
