Instagram chief admits Meta hid low teen use of safety tools

Adam Mosseri, who oversees Instagram at Meta, admitted in a California court on Tuesday that he touted newly launched safety tools for teens without disclosing low adoption rates from early testing several years ago.

With more than three billion users around the world, Meta is fending off accusations by a coalition of 29 US states that it deliberately designed its products to hook children and gather their data while misleading the public.

The trial entered its second week on Monday and is expected to last through late September.

A lawyer with the Colorado Attorney General's office, Jason Slothouber, questioned Mosseri about the low usage of the tools, known as Take a Break and Quiet Mode, which were launched in 2021 and 2023, respectively.

Other witnesses have testified over the past week that these tools weren't effective, though Mosseri pushed back on the idea that adoption remains low.

"This is one feature with a disappointingly low usage rate which since this time has significantly improved," Mosseri said.

On December 8, 2021, Mosseri testified at a Congressional hearing about the negative impacts of social media on young people. One day before, Meta published a blog post, authored by Mosseri, about rolling out new safety features for teens.

"Early test results show that once teens set the reminders, more than 90% of them keep them on," the blog post says.

However, that number referred to the number of people who kept it on after turning it on.

When asked if Meta disclosed its low adoption, which only reached 1 or 2 percent of the accounts, Mosseri admitted the company did not but said the safety features were later turned on by default in Teen Accounts, which have parental controls and launched in 2024.

However, Mosseri couldn't recall if Meta had ever disclosed the percentage of teen accounts that have parental controls turned on.

Mosseri will continue testifying on Wednesday. Founder and CEO Mark Zuckerberg is also expected to testify, though it's unclear when.