![\"\"](\"https:\/\/world.edu\/wp-content\/uploads\/2024\/07\/microsoft-access.jpg\")
<\/p>\n<\/p>\n
In May, a multibillion-dollar UK-based multinational called Informa announced in a\u00a0trading update<\/a>\u00a0that it had signed a deal with Microsoft involving \u201caccess to advanced learning content and data, and a partnership to explore AI expert applications\u201d. Informa is the parent company of\u00a0Taylor & Francis<\/a>, which publishes a wide range of academic and technical books and journals, so the data in question may include the content of these books and journals.<\/p>\n According to\u00a0reports published last week<\/a>, the authors of the content do not appear to have been asked or even informed about the deal. What\u2019s more, they say they had no opportunity to opt out of the deal, and will not see any money from it.<\/p>\n Academics are only the latest of several groups of what we might call content creators to take umbrage at having their work ingested by the generative AI models currently racing to hoover up the products of human culture.\u00a0Newspapers<\/a>,\u00a0visual artists<\/a>\u00a0and\u00a0record labels<\/a>\u00a0are already taking AI companies to court.<\/p>\n While it\u2019s unclear how Informa will react to the rumblings of discontent, the deal is a reminder to authors to be aware of the contractual terms of the publishing agreements they sign.<\/p>\n Informa\u2019s update stated four focus areas of the Microsoft deal:<\/p>\n Informa will be paid more than \u00a38 million (A$15.5 million) for initial access to the data, followed by recurring payments of an unspecified amount for the next three years.<\/p>\n We don\u2019t know exactly what Microsoft plans to do with its data access, but a likely scenario is that the content of academic books and articles would be added to the training data of ChatGPT-like generative AI models. In principle this should make the output of the AI systems more accurate, though existing AI models have faced heavy criticism, not only for\u00a0regurgitating training data<\/a>\u00a0without citation (which can be viewed as a kind of\u00a0plagiarism<\/a>), but also for\u00a0inventing false information<\/a>\u00a0and\u00a0attributing<\/a>\u00a0it to real sources.<\/p>\n However, the update also says \u201cthe agreement protects intellectual property rights, including limits on verbatim text extracts and alignment on the importance of detailed citation references\u201d.<\/p>\n The \u201climits on verbatim text extracts\u201d mentioned likely pertains to the\u00a0US doctrine of fair use<\/a>, which permits certain uses of copyright-protected material.<\/p>\n Many generative AI companies are currently\u00a0facing copyright infringement lawsuits<\/a>\u00a0over their use of training data, and their defences are likely to rely on claiming fair use.<\/p>\n The \u201cimportance of detailed citation references\u201d may pertain to the concept of attribution in copyright. This is a\u00a0moral right<\/a>\u00a0possessed by authors. It provides that the creator of the work should be known and attributed as the author when their work is reproduced.<\/p>\n Most academics do not receive payment or make any profit from most of their scholarly publishing. Rather, writing journal and conference papers is usually considered part of the scope of work within a full-time, tenured position. Publication builds an academic\u2019s credibility and promotes their research.<\/p>\n The basic process often goes like this: an author researches and writes an original article and submits it to a journal publisher for peer review. Most peer reviewers and editorial board members also receive no payment for their work.<\/p>\n In fact, some journals may require authors to pay an \u201carticle processing charge<\/a>\u201d to cover editing and other costs. This can be thousands of dollars for an\u00a0open access<\/a>\u00a0publication. Generally speaking, the more prestigious the publication, the higher the charge. Just found out that Taylor & Francis has sold access to all @routledgebooks<\/a> data to Microsoft to train their AI. This includes my publications. I get no payment for using my research labour, never mind the bigger problems of this energy-intensive extractivism…<\/p>\n — Dr Ruth Alison Clemens (@RuthieClems) July 17, 2024<\/a><\/p><\/blockquote>\n What\u2019s in the Informa deal?<\/strong><\/h4>\n
\n
How does scholarly publishing usually work?<\/strong><\/h4>\n
\n\n