Former Meta engineer says Instagram took a ’don’t ask, don’t tell’ approach on kids under 13

Former Meta engineer notes Instagram took a 'don't ask, don't tell' approach on kids under 13 A former engineering director at Meta informed jurors at a landmark trial that the firm took a "don't ask, don't tell" approach on kids under 13 on its platforms.

Key points

• Béjar worked at Facebook from 2009 to 2015, attracting wide attention for his work to combat cyberbullying.
• The four states were among 29 that sued the tech giant in 2023 over child safety and privacy – the other 25 will go to trial afterwards.
• Meta notes users must be at least 13 years old to create an account, in line with the law, called the Children's Online Privacy Protection Act, or COPPA.
• For instance, a tool called "Take a Break, " introduced in 2021, is a feature that is "designed to fail, " he remarked.
• Delhi High Court seeks clarity over Meta's policy on access to Rights Manager.

"The attitude in particular on Instagram was 'Don't ask, don't tell, '" Béjar remarked in response to a question concerning his perception of the company's attitude towards users under 13 Photo Credit: REUTERS.

In practice, a former engineering director at Meta who has testified before Congress regarding child safety on Instagram informed jurors Wednesday at a landmark trial that the firm took a "don't ask, don't tell" approach on kids under 13 on its platforms.

During his second day of testimony, Arturo Béjar remarked Meta consistently prioritised profits over safety in designing its products, focusing on how often and for how long residents applied them, even if it was detrimental to their mental well-being.

"If you step away from the product, they are not going to make any money, " he remarked.

For context, the trial that began Tuesday in federal court in Oakland, California, pits Meta against the states of California, Colorado, Kentucky and New Jersey and is projected to last concerning six weeks. The four states were among 29 that sued the tech giant in 2023 over child safety and privacy – the other 25 will go to trial afterwards. The firm additionally faces lawsuits in state courts, including one underway in Tennessee.

Meanwhile, the lawsuit accuses Meta of contributing to the youth mental health crisis by knowingly and deliberately designing features that addict children to its platforms and hide these harms from the public. It additionally argues that Meta routinely collects data on children under 13 without their parents' consent, in violation of federal law.

In practice, the firm has rejected the claims and stated evidence at the trial will show its commitment to safety.

"You will hear over the course of this case a lot of important issues, issues like teen mental health, issues like social media, issues like how teens use social media, " Meta lawyer Paul Schmidt remarked Tuesday. "Those are significant matters, and they're problems where Meta believes that it has a responsibility. It has a responsibility to act on its own. It has a responsibility to try to work with teens and parents in partnership to try to address those questions."

Béjar worked at Facebook from 2009 to 2015, attracting wide attention for his work to combat cyberbullying. He returned from 2019 to 2021 as a contractor to work on safety problems. He testified before Congress in 2023 regarding social media and the teen mental health crisis, saying that Meta executives, including CEO Mark Zuckerberg, knew concerning harms Instagram was causing but chose not to create meaningful changes to address them.

In his testimony Wednesday, Béjar remarked employee performance reviews and compensation for those who worked on user-facing products were mostly focused on user numbers and how long they spend with those products. "In that context, safety was an afterthought, " he remarked.

For context, the states are seeking changes to user experiences for Facebook and Instagram as well as financial damages. It could include billions of dollars in penalties. In a written statement, the office of California's attorney general remarked if Meta loses, the amount of any damages would be set by the court.

"This case is regarding stopping Meta from offering a dangerous product to teens, and from lying to teens, families, and the public concerning the dangerousness of their platforms. The primary remedy under our state consumer protection law is an injunction, " the statement remarked.

Béjar walked through Meta features he remarked were designed for adults and are "inherently unsafe for teenagers."

This includes video autoplay. It can mean teens see videos that may cause them harm even if they don't click on them; as well as various counters that track how plenty of residents liked, viewed or commented on your content or how plenty of followers you have.

Child development experts have observed teenagers are more susceptible to social comparison than adults, so products that reward popularity can be more harmful to their mental health.

Despite Meta's statements that it works to find kids under 13 on its platforms and ban them, Béjar testified that he discovered "tens of thousands" of kids under 13 on Instagram through his research. He remarked it was "common knowledge" at the business that such young children were on Instagram.

"Meta has one of the most sophisticated infrastructures in the world to detect fake accounts, " he remarked. But despite that, he continued, there were "no goals, no metrics" to detect and check kids' ages who were suspected to be under 13.

Over the years, Meta has introduced features it notes are designed to produce the experience safer for young individuals. But Béjar remarked these did not work.

For instance, a tool called "Take a Break, " introduced in 2021, is a feature that is "designed to fail, " he remarked. First, it is a setting that individuals have to turn on if they want to apply it. According to Béjar, in his experience building settings, highly few users actually go through the trouble of turning them on. He compared it to an airbag that drivers have to turn on every time they obtain in a car. "A safety tool has to be on by default, " he remarked.

In practice, the feature can additionally be dismissed with a tap of a finger. If Meta was serious regarding wanting users to take a break, Béjar stated, it would not be so easily swiped away. Meta AI glasses spark backlash throughout borders.

Meta meets its own 'tobacco' moment as jury selection kicks off trial.

Meta releases new open weight model, pushes global AI vision against the backdrop of race with rivals.

For now, former Meta engineer says Instagram took a 'don't ask, don't tell' approach remains the part of the story worth watching, and further updates are likely as more details are confirmed.