Meta has made two AI-driven features more widely available for advertisers, with the aim of providing them with more brand safety, transparency and control over the proximity of their ads to suitable content.
In Meta’s ad manager and brand safety hubs, advertisers will be able to select one of three inventory filters that controls the kind of content that appears above and below an ad on Facebook and Instagram feeds, using AI detection.
“Expanded inventory” is the default setting advertisers would be used to, showing content that adheres to Meta’s Community Standards and meets its monetization eligibility criteria. The “moderate” inventory filter eliminates high-risk content that, broadly speaking, depicts things like tragedy, conflict, controversial social issues, substance abuse, sexual subjects or crime, but will include content where those subjects are discussed. The “limited” inventory filter is the most conservative and excludes content that may be considered high and medium risk, only allowing for content where potentially sensitive subjects are discussed or referenced, but are not the focus.
The new controls align with GARM’s (Global Alliance for Responsible Media) Suitability Framework, and the company has released a guide as to what is and is not included in each filter level.
“As a company, we’ve really been listening to the needs of the advertising community, which are essentially that they would like to have greater control and greater transparency in the context of the content where their ads are appearing,” says Samantha Stetson, VP, client council and industry trade relations at Meta. “What is really important to know is this is in a very much collaborative industry initiative.”
In addition, Meta has introduced third-party brand suitability verification on Facebook Feed. This is an independent AI-powered solution created along with Zefr, a brand suitability company. Zefr’s AI product assesses video, image, test, and audio and labels content on Facebook feeds, also based on GARM suitability standards.
This will allow advertisers to verify the suitability of content their ads appear next to, as well as measure results and make more informed brand safety decisions.
Over the last several months, Meta has been testing the two new features with 25 advertisers.
Stetson says that one of the biggest evolutions that came out of the creation of the new features was the really clear distinction between what is considered brand safety, and what is considered brand suitability.
“When you think about brand safety, that is the content that no advertiser wants to be associated with and does not want monetized. Frankly, it’s the content that we don’t want on our platform, either. It is not good for people. Brand safety is essentially people safety in my mind and GARM established the brand safety floor, which is aligned to our Community Standards and Guidelines.”
The inventory filters for Facebook and Instagram Feeds are now rolling out to advertisers in English and Spanish-speaking markets, with additional languages and markets added throughout the year. The third-party verification solution for is also now available through Zefr, with additional Meta Business Partners will be added in the coming months.
