The watchdog is attacking Meta's “incoherent” policy on deceptively manipulated videos, saying it is too focused on artificial intelligence.
Joe Biden, US President PHOTO: Archive
Meta's supervisory board has found that a Facebook video falsely suggesting that US President Joe Biden is a pedophile does not violate the company's current rules, while deeming those rules to be “incoherent” and focuses too little on content generated by artificial intelligence, writes theguardian.com.
The board, which is funded by Facebook's parent company Meta but operates independently, took up the Biden video case in October in response to a user complaint about an altered seven-second video of the president.
It ruled that Meta was right to let the video go under its current policy, which bans deceptively altered videos only if they were produced by artificial intelligence or make people appear to be saying words they didn't never said
But the ruling is the first to criticize Meta's policy regarding
“manipulated environments”amid growing concerns about the possible use of new artificial intelligence technologies to influence this year's elections.
According to her, the policy “lacks compelling justification, is inconsistent and confusing for users, and does not clearly specify the harms it seeks to preventHe suggested that Meta update it to cover both audio and video content and apply tags to identify that it has been manipulated, regardless of whether artificial intelligence was used or not.
He did not request that the policy also apply to photos, as that would be too difficult to enforce at the Meta level.
Meta, which also owns Instagram and WhatsApp, informed the board during the review that it plans to update the policy “to respond to the evolution of a new and increasingly realistic AI”according to the decision.
The Facebook clip manipulated real footage of Biden exchanging stickers “I Voted”
with his granddaughter during the 2022 US midterm elections and kissing her on the cheek.
The board stated that content modified without the aid of AI is “widespread and not necessarily less deceptive” than content generated by AI tools.
The council added that enforcement should consist of labeling content, rather than Meta's current approach of removing posts from its platforms.
The company said in a statement Monday that it is reviewing the ruling and will respond publicly within 60 days.