Meta will identify AI-generated content on Facebook, Instagram and Threads

Meta announced last Friday (5) that it would Tag AI-generated content posted by users Facebook, instagram This Thread. The measure, which will come into force from May, follows a recommendation from the tech giant’s Supervisory Board that updated its manipulated media policy established in 2020.

According to the company, Voices, photos and videos created by AI-based tools will receive tags. Identification of these materials will occur if users report using the technology or if the social networks detect “artificial intelligence image indicators that comply with industry standards.”

The operation of this detection system was not detailed by the company, which stated that it had it. Network of nearly 100 independent fact-checkers. They are responsible for reviewing “false and misleading content” generated by artificial intelligence and are working to combat misinformation on the holding’s social networks.

With the change, AI-generated content will receive the “Made with IA” or “Made with IA” seal, with free translation. If Meta considers material to have greater potential to mislead the public about a relevant topic, the tag may be accompanied by information to contextualize it.

Removal of AI-generated content

Updating Meta’s policies on manipulated content will prevent the restriction of such material posted on Instagram, Facebook and Threads as is currently the case. As mentioned above, the company can add context to specific posts to educate users if necessary.

According to Big Tech, this eliminates “the risk of unnecessary restriction of freedom of expression.” It’s worth noting that the company is already labeling photorealistic images created with its own AI tool and adding the “Imagined with AI” seal.

However, Monika Bickert, Meta’s vice president of content policy, emphasized: No changes have been made to removing content that violates community guidelines. In other words, material may be deleted if it does not meet standards for intimidation, election interference, and harassment, regardless of whether it is created by humans or bots.