Politicians’ Books Illegally Copied for Meta AI Training
Allegations Against Meta: The Use of Pirated Works in AI Development
Recent revelations have surfaced involving allegations that Meta, the parent company of Facebook, has utilized a collection of pirated digital materials to enhance its artificial intelligence software. This information raises concerns about copyright violations and the ethical implications of using unauthorised content.
The Controversial Library Genesis
Library Genesis (often referred to as LibGen) is recognized as a significant source of pirated books and articles. Among the contents exposed are works by high-profile politicians and authors, including articles by the Prime Minister and books written by political figures such as Rachel Reeves, Lisa Nandy, and Sir Chris Bryant. For instance, Reeves’s biographies of female economists and Nandy’s proposals to improve Britain are included in this notorious archive.
Legal Troubles for Meta
Meta is currently facing legal challenges in the United States, accused of infringing copyright laws by using works from LibGen for training its AI systems. The company, however, staunchly denies these allegations. The timing of these events is particularly interesting as the government is advocating for reforms in copyright laws, which would allow AI companies to use copyrighted material without explicit permission unless the owner opts out. Critics within the creative industry have condemned this proposed change, labeling it as a substantial theft of intellectual property.
Creative Industry’s Response
Many authors and artists are expressing outrage over the possible implications of such copyright law reforms. Notably, Sir Keir Starmer, Rachel Reeves, Lisa Nandy, and Sir Chris Bryant—key figures in the proposed reforms—could find themselves in a predicament, as they are also potential victims of this alleged infringement.
Starmer’s academic articles, including those in legal journals, are also part of the LibGen collection. The creative industry remains unconvinced that opting out will be effective, alleging that there is no practical method for ensuring protection against unauthorized use.
Discoveries and More Allegations
The Atlantic magazine recently unveiled a searchable version of LibGen, enabling authors to check whether their works have been pirated. This move has led to significant discoveries; for instance, thriller author Brad Thor found 126 of his books within LibGen, leading him to condemn Meta’s actions on social media.
Additionally, the dataset in question includes a book co-authored by Matt Clifford, the government’s chief AI adviser. This connection underscores the complexities involved when prominent figures in the government are linked to works found in a library infamous for copyright violations.
Ethics and Internal Discussions
In light of these developments, internal communications at Meta have come under scrutiny. Documents from the ongoing lawsuit highlight conversations among Meta employees about the ethical ramifications of using material from LibGen. Certain employees have referred to LibGen as an "illegal pirated website," expressing their concerns about incorporating such resources into the development of AI models.
Notable comments include one engineer questioning the legality of using content from LibGen, and another mentioning the potential legal ramifications. This evidence raises the question of how deeply the ethical considerations were weighed before decisions were made concerning the use of pirated content.
Extensive Piracy Allegations
The scope of the alleged piracy is expansive. Legal filings indicate that Meta engineers employed torrenting protocols to expedite the downloading of the pirated works. This decision, it is contended, transformed Meta into a "host and distributor" of over 267 terabytes of pirated materials—approximately equivalent to 20 times the size of the entire print collection at the Library of Congress.
Moreover, the legal brief indicates that works authored by Supreme Court justices are also included in the pirated collection. The legal stance taken by Meta claims that its actions fall under "fair use" as defined by U.S. copyright law, though many experts dispute this justification.
Notable Works Found in LibGen
Some of the titles identified in LibGen include:
- Alice in Westminster: The Political Life of Alice Bacon by Rachel Reeves
- Women of Westminster: The MPs Who Changed Politics by Rachel Reeves
- James and John: A True Story of Prejudice and Murder by Sir Chris Bryant
- All In: How We Build a Country That Works by Lisa Nandy
- How to Be a Founder by Alice Bentinck and Matt Clifford
Additionally, Starmer has articles featured in esteemed journals discussing topics like human rights and domestic violence.
These events prompt a deeper examination of the intersection between technology and copyright, highlighting a growing tension within the creative industries and the implications for both artists and tech companies alike.
