Meta Expands AI Labeling Policies As 2024 Presidential Race Approaches

Meta will develop It is labeling of artificial intelligence-generated content in the middle of widespread concern that A increase In “deepfake” posts will deceive social media users.

THE business said Friday that he will apply "Do with AI" Labels has A range of video, audio And pictures. The meta previous Strategies against manipulated media, which were writing In 2020 Before THE technology industry AI boom, only applied has video content.

THE expansion follows critical Since The meta Monitoring Advice, which called THE the company existing Strategies "incoherent" In FEBRUARY After A amended video of President Biden stayed on Facebook because he doesn't violate The meta rules. THE Monitoring Advice, A out band finance by Meta, called on THE business has extend THE policy has address amended audio as GOOD as videos When they falsely represent people TO DO things.

Advertisement

"We accept with THE Monitoring Boards argument that OUR existing approach East Also narrow" because he only applies has fake speech And not amended Actions, Meta Vice President of Content Policy Monique Bickert said In A declaration.

"OUR manipulated media policy was writing In 2020 When realistic AI-powered content was rare And THE overall concern was about videos. »

Following THE Monitoring Boards recommendation, Meta Also agreed has No longer withdraw numerically created media if he doesn't violate any of them other rules, but THE business will attach A label saying that THE content has has been amended. Departure following month, THE business will to start has apply "Do with AI" Labels on content he determined East AI Or When people disclose they are Download AI-powered content.

In FEBRUARY, Meta revealed plans has develop A system has identify AI-powered content that users create using services Since other technology companies which to have agreed has to integrate A AI identifier Or A watermark.