Meta, the parent company of Facebook, will be adjusting its policies on manipulated and A.I.-generated content ahead of the fall elections. This decision comes after an independent body overseeing the company’s content moderation found that the previous policies were “incoherent and confusing” and needed to be reconsidered. The changes were prompted by the Meta Oversight Board’s recommendations earlier this year, following a review of a highly edited video of President Biden that was circulating on Facebook. The video had been manipulated to make it appear as if Mr. Biden was inappropriately touching his adult granddaughter’s chest, creating a misleading impression.
The Oversight Board determined that the video did not violate Meta’s policies because it had not been manipulated with artificial intelligence (AI) and did not show Mr. Biden saying or doing anything he did not actually say or do. However, the board criticized Meta’s current policy as lacking justification and focusing too much on how content is created rather than preventing specific harms, such as disrupting electoral processes. In response, Meta Vice President of Content Policy Monika Bickert announced that the company would start labeling AI-generated content in May and adjust its policies to label manipulated media with informational labels and context, rather than simply removing the content based on community standards violations.
The new labeling approach for manipulated content will cover a broader range of content beyond what the Oversight Board had recommended. If Meta determines that digitally-created or altered images, videos, or audio pose a high risk of materially deceiving the public on important matters, the company may add more prominent labels to provide additional information and context. Meta acknowledged that its previous policy on manipulated videos had been too narrow, focusing on AI-generated content that makes a person appear to say something they did not say. With advancements in AI technology, concerns about manipulation through audio and photos have also become important to address.
The Oversight Board welcomed Meta’s commitments to changes in how the company handles manipulated content, recognizing the significance of these adjustments. This decision comes at a time when AI and editing tools make it easier for users to create realistic-seeming video and audio clips, raising concerns about misinformation and voter suppression. Instances of AI-generated content related to former President Trump and President Biden continue to be shared online, highlighting the need for platforms like Meta to establish clear and effective policies for identifying and addressing manipulated content. As the fall elections approach, it will be important for Meta and other social media companies to prioritize transparency and integrity in their content moderation practices.
