On Wednesday, Meta Platforms faced backlash over its use of pirated books for AI training and content moderation service, renewing scrutiny over any previous Meta copyright violations and practices.
The Meta AI battle is intensifying the giant’s legal challenges for using pirated content to train its AI models. According to Reuters, internal communications revealed that Zuckerberg, Meta Chief, personally approved the use of data from Library Genesis (LibGen) – known for hosting pirated books – to develop the company’s Llama AI models.
Zuckerberg’s decision to make that move will go as one of Meta’s many AI copyright violations.
The Meta AI battle, for a long time caused bumps from two big issues that could redefine the public perception of the Big Tech giant, with reports of using pirated material to train AI models have increased creators and debates about the rights in the age of AI copyright violations.
AI Training Pirated Books
According to Wired, Meta used copyright-protected books to train its AI models through torrent site Library Genesis. According to the court documents from the ongoing copyright lawsuit, employees at Meta used that material to build datasets for its AI systems.
One Meta AI training engineer highlighted that downloading from a corporate device was against the rules and escalated the issue to CEO Mark Zuckerberg, who is referred to as “MZ”. The revealed documents show that Zuckerberg approved the use of pirated materials for training AI.
“These newly unredacted documents reveal exchanges between Meta employees unearthed in the discovery process,” Wired noted, revealing that this issue will be another blow to Meta’s reputation, among creators whose rights may have been violated.
It’s not a surprise that Meta would do something like this, it has a long history of ethical scandals, including the issue with Cambridge Analytica and the breaches of user privacy. Critics say this new Meta copyright case is just one more example of how Meta prefers competitive advantage to use friendly and legal behavior.
A Shift to “Community Notes”
On January 7, Zuckerberg announced it will discontinue its US fact-checking program, by replacing it with an X-lookalike community-driven system, similar to Musk’s microblogging platform.
The “Community Notes” feature will allow users to collaborate to add contextual edits or corrections to any misleading post. Some are viewing this as a step toward free speech, critics however disagree, labeling the change as a window for the spread of misinformation.
While “Community Notes” will apply to organic posts for now, paid advertisements and branded content will be exempted, at least for the time being.
“We are making the transition to Community Notes over the next couple of months in the United States and, as with any new product rollout, we’ll continually evaluate and improve it over the course of the year,” said a Meta spokesperson.
The timing of the Meta copyright policy shift come just in time for the elected President Donald Trump’s inauguration.
Zuckerberg’s Leadership Skills
Both arguments point to one important constant in Meta’s operations, which has become a business focused platform rather than a platform that leads and aids in ethics. Training AI data sourcing with pirated material and changing the moderation system according to political decisions signal what is at risk in Meta maintaining its leading positions which is leading to the loss of the users’ trust.
With growing attacks on the platform, these over piling issues may have far reaching implications not only for the Meta legal battle reputation but also for how the wider tech industry approaches ethical AI and takes responsibility for its Meta copyright content.
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Tech sections to stay informed and up-to-date with our daily articles.