Meta will add a Made with AI label to AI-generated content starting May

Starting in May this year, Meta will add a Made with AI label to more AI-generated content. It will broadly label AI-generated media including videos, images, and audio on all its platforms including Facebook, Instagram, and Threads.

Let’s not forget that in February, Meta hinted to expand the labeling of AI content shared on its platforms. However, the announcement and detailed plans were shared by the VP of the company’s Content Policy on Friday, April 5.

Meta’s decision to label AI content comes after the Oversight Board’s feedback

Monika Bickert, VP of Content Policy at Meta recently informed about the company’s plans for AI content labeling. Bickert mentioned that the company will start labeling manipulated content when the industry standard AI image indicators hint at it. This implies further to the AI-generated content uploaded by any user disclosing its nature.

According to the blog post, Meta plans to label and contextualize the manipulated media rather than remove it from its platform, risking people’s freedom of speech. Meta agrees with the Oversight Board’s feedback about the company’s narrow labeling approach, which only covered manipulated videos. 

Adding more on the Meta’s manipulated media policy, Bickert wrote: 

Our manipulated media policy was written in 2020 when realistic AI-generated content was rare and the overarching concern was about videos. 

The decision to expand AI-content labeling comes after feedback from the Oversight Board in the wake of a controversial manipulated video of President Biden inappropriately touching her granddaughter.

Worth noting that the Oversight Board earlier agreed to Meta’s decision to not remove the controversial video from Facebook. The company then insisted that the video didn’t violate their manipulated media policy. However, the board asked to reconsider this policy quickly, given the number of elections in 2024.  

The company plans to widen its labeling approach starting May 2024. But, it will stop removing content that violates the manipulated video policy from July. Meta says that this will allow people to understand more about the self-disclosure process before it stops removing manipulated media in batches.

All in all, it won’t be wrong to say that Meta’s plan to label AI content is a measure to mitigate misinformation given the upcoming US elections. Time will tell how well it implements the said changes.