Instagram will begin scanning messages sent and received to protect people under 19 from “inappropriate images,” Meta announced.
The feature is being kept secret until “later this year” and will also work with encrypted messages, a spokesperson said, indicating that the company intends to implement a so-called client-side scanning service for the first time.
However, the update does not address the controversial request to report inappropriate messages to Instagram's servers.
Instead, only a user's personal device will know whether a message has been filtered, a promise that has led to criticism as another example of companies “grading their own homework.”
“We plan to release new features aimed at preventing teens from seeing unwanted and potentially inappropriate images in messages from people they're already connected to.” says the company. said in a blog post“And to discourage people from sending these kinds of images themselves. We'll be rolling out more on this feature later this year, which also works with encrypted chats.”
This is Meta's first proposal for how it plans to respond to criticism that its plans to encrypt direct messages on Facebook Messenger and Instagram could put children and young people at risk.
A rough description of the desired functionality is as follows: “Communication safety” setting The feature, introduced by Apple in 2023, detects nude photos and videos sent to a child's device and automatically blurs them, but requires the child to view them or contact a trusted adult. provide options.
The plan falls short of the enhanced client-side scanning required by child safety groups to ensure such inappropriate messages are more proactively reported to the service's moderators and to track and catch repeat offenders. You will be able to do this.
The move comes as Mehta faces a lawsuit in New Mexico for failing to protect children on its platform. Meta estimates that about 100,000 children on Facebook and Instagram are sexually harassed online every day, according to unsealed legal filings related to the case last week. .
On Wednesday, Meta CEO Mark Zuckerberg will appear before the US Congress along with other social media executives for a hearing on child safety.
Instagram announced small, immediate updates to teen safety features on its platform, along with promising future scanning tools. Users under 19 will default to a privacy setting that prevents people who don't follow them from sending them direct messages. Previously, this restriction only applied to adults sending messages to teens.
After newsletter promotion
Other new features are also being released for parents who use the service's “monitoring” tools. This will allow parents to connect their Instagram account to their child's account to set time limits, monitor their teen's blocks, and receive notifications when settings change. become. Parents will now be required to actively approve or deny attempts by children under 16 to lower safety settings.
“As with all of our parental monitoring tools, this new feature is designed to help parents and teens navigate their online lives together and help them decide what's best for them and their families offline. The idea is to encourage conversation,” Mehta said.
Arturo Béjar, a former senior engineer and consultant at Meta, said the changes should be accompanied by regular updates about how many unwanted advances teenagers have received on Instagram. Ta. Without such data, there is no way to assess the impact of safety updates. According to an independent survey of Instagram users conducted by Bejart in 2021, one in eight children aged 13 to 15 who use Instagram have experienced unwanted sexual advances. .
“This is also a promise that we will grade our own homework,” he says. “Until they start reporting quarterly unwanted advances, as teens are experiencing, how will we know if they kept their word or what impact it had? Even after more than two years of knowing that 1 in 8 kids receive unwanted advances every week, there's still no way for teens to flag or report unwanted advances. ”
