Meta on Tuesday announced it would now include information on how it uses machine learning to deliver ads — increasing transparency since Meta and others including Google have launched independent tools to tell users why they see the ads they do. 

The technology for Meta’s updated version of its “Why am I seeing this ad?” tool will provide information on how someone’s activity — on and off its platform such as Facebook and Instagram — inform the company’s machine learning models.

It also includes new examples and illustrations to explain how these models connect various topics to show relevant ads, according to Pedro Pavón, global director of ads and monetization privacy at Meta.

“We worked closely with external privacy experts and policy stakeholders from around the world to get input on what transparency changes they want to see in our ads system,” Pavón wrote in a post. “A consistent answer was that we should increase our transparency around how our machine learning models contribute to the ads people see on our services.

Making it easier for people to find Meta’s ads controls, marketers will be able to access Ads Preferences from additional pages in the “Why am I seeing this ad?” tool.

Today’s update takes it one step further to include information about how Meta uses machine learning models to show people ads.

Starting today, Meta has summarized information into topics about how the activity on and off its technologies — such as liking a post on a friend’s Facebook page or interacting with a favorite sports website — may inform the machine-learning models used to shape and deliver the ads people see. 

Meta has also included new examples and illustrations explaining how machine-learning models connect various topics to show relevant ads.  

Pavón explains that by stepping up Meta’s transparency around how its machine-learning models works to deliver ads, the company aims to help people feel more secure and increase our accountability.

“The changes we’re making to “Why am I seeing this ad?” reflect the feedback we’ve received and are designed to provide people with clear information about the machine learning models that help determine the ads they see,” Pavón wrote.