Facebook to provide additional insights on how machine learning determines the ads you view

Some updates are being made to the 'Why am I seeing this ad' tool.

Meta is enhancing Facebook's ad transparency features to provide a clearer understanding of how it utilizes machine learning to determine the ads you receive. As of Tuesday, when you select the "Why am I seeing this ad?" option on an ad, you will be presented with the following details, as outlined in a blog post by Meta's Pedro Pavón:

Information summarized into topics about how your activity both on and off our technologies — such as liking a post on a friend’s Facebook page or interacting with your favorite sports website — may inform the machine learning models we use to shape and deliver the ads you see.

New examples and illustrations explaining how our machine learning models connect various topics to show you relevant ads.  

More ways to find our ads controls. You will now be able to access Ads Preferences from additional pages in the “Why am I seeing this ad?” tool.

Pavón notes that external privacy experts and policy stakeholders have recommended that the company enhance its transparency regarding how machine learning is utilized to select which ads are displayed. According to Pavón, the company is dedicated to the responsible use of machine learning models. It is vital to be transparent about the utilization of machine learning, as it ensures that individuals are aware of the role of this technology in the advertising system and the type of information it employs, Pavón added. Although the new information was not observed in any ads on their own Facebook feed, a few screenshots available on Pavón's blog can provide an idea of what to anticipate.

Initially, the enhanced messaging will be introduced on Facebook, with plans to extend it to Instagram "at a later date." TikTok has also recently become more forthcoming about the factors that influence the content presented in your For You feeds by implementing a comparable feature called "Why this video" in December. This feature provides information about the specifics of the content you are viewing.