by Katie Paul
NEW YORK (Reuters) – Facebook owner Meta Platforms announced significant changes to its rules on digitally created and transformed content on Friday, ahead of a U.S. election that is expected to test its ability to fight against disinformation in the face of the development of artificial intelligence (AI).
The social media giant will begin putting “Made with AI” labels on artificial intelligence-generated videos, images and sounds posted on its platforms in May.
Meta is thus extending its rules, which were until now limited to certain modified videos, underlined in a blog Monika Bickert, vice-president of the group responsible for content rules.
Monika Bickert added that Meta would also place separate, more visible labels on digitally altered content that poses a “particularly high risk of significantly misleading the public on an important topic”, regardless of whether the content was created using AI or other tools.
This new approach will change the way the American group handles modified content. Meta will move from an approach based on deleting a limited number of messages to a strategy of maintaining content while providing Internet users with information on how it was created, explains the group.
Meta has already announced a project to detect images created using generative AI tools from other companies using invisible markers embedded in files, without specifying the timetable for the deployment of this service.
A company spokesperson told Reuters that the new labeling approach would apply to content published on Facebook, Instagram and Threads, three of the group’s services. WhatsApp messaging and Quest virtual reality headsets, however, are not affected by the new rules.
(Reporting Katie Paul; French version Alban Kacher, edited by Claude Chendjou)
Copyright © 2024 Thomson Reuters
Are you following this action?
Receive all the information on META in real time: