Meta’s oversight board determined that a Facebook video falsely alleging that U.S. President Joe Biden is a pedophile did not violate the company’s current rules, but expressed that the rules were “disjointed”. It was acknowledged that the focus is too narrow on AI-generated content.
The board, which is funded by Facebook’s parent company Meta but operates independently, took on the Biden video case in October after receiving user complaints about a doctored seven-second video of the president.
The board ruled that under current policies, the misleading altered video would only be prohibited if it was created by artificial intelligence or made to appear to say words that were not actually said. Therefore, Meta was correct in continuing to publish the video.
This ruling is the first to criticize Meta’s policies against “manipulated media” amidst concerns about the potential use of new AI technology to influence upcoming elections.
The board stated that the policy “lacks a convincing justification, is disjointed and confusing to users, and does not clearly articulate the harms it seeks to prevent.” It suggested updating the policy to cover both audio and video content, and to apply a label indicating that it has been manipulated, regardless of whether AI is used.
It did not require the policy to apply to photos, as doing so could make enforcement too difficult at Meta’s scale.
Meta, which also owns Instagram and WhatsApp, informed the board that it plans to update its policies to address new and increasingly realistic advances in AI, according to the ruling.
The video on Facebook is a manipulated version of real footage of Biden exchanging “I voted” stickers with his granddaughter and kissing her on the cheek during the 2022 US midterm elections.
The board noted that non-AI modified content is “more prevalent and not necessarily less misleading” than content generated by AI tools.
It recommended that enforcement should involve applying labels to content, rather than Meta’s current approach of removing posts from the platform.
The company announced that it is reviewing the ruling and will respond publicly within 60 days.
Source: www.theguardian.com
