Parenting has and always will come with a steep learning curve. Where questions are sometimes basic such as How can I get my child to eat? or Why is my teen so moody?, today’s parents also face things like How can I prevent my teen from sexting?. As issues like sending or receiving sexually explicit material grow more pervasive among younger people — both in terms of participation and harassment — private entities and public corporations alike have been working on solutions. One that’s showing promise is in artificial intelligence (AI).
Using AI to Filter Sexual Content
Though the term sexting refers to the act of sending sexually explicit content or images via text message, more teens are beginning to use direct messaging on social media sites such as Instagram, Facebook and SnapChat to exchange intimate photos and words. What are these social media sites doing to moderate content? Unfortunately, the answer is, not much.
It’s not that social media sites aren’t concerned about teen sexting and unsolicited sexual advances — they are —these companies have too many variables to worry about. Though sites such as Twitter claim that in order to create a program that is capable of identifying all forms of offensive content — be it an erect human penis, voyeuristic snapshots or suggestive text — would be “complex,” the truth is that such technology already exists (as evident by the apps discussed below).
The real issues lie with censorship itself, which social media companies are loath to implement. This is for a couple different reasons.
First, not all offensive images and content are unsolicited. Although many associate private images with harassment, there are certain users who consent to receive them. If a system flags these images as violations and punishes both the sender and recipient, would users turn to another platform?
Second, if a program existed to identify and filter sexual content, would the sites apply it only to teen users, or would it be a sitewide filter? If it was sitewide, would users revolt and claim too aggressive of censorship? If it applied only to teens, what’s to say teens wouldn’t run out and create a new account with a false birthdate?
At the end of last year, Tumblr instituted a NSFW ban. Not only did it start blocking images from legitimate (non-pornographic) blogs, but then it missed some areas with adult images and content along the way. This has resulted in many users leaving the platform and possibly harming Tumblr’s longterm existence. While one could argue their AI wasn’t good, it also shows the difficulty of instituting this kind of tech on the social media landscape.
Third-Party Apps May Be Parents’ Best Bet
Though social media isn’t using AI to screen for sexual content (as of yet), third-party apps have no qualms about helping parents censor inappropriate images. For instance, Bumble’s Private Detector feature uses AI to identify and blur out adult content and boasts an impressive 98% accuracy rate. Though Bumble does not remove explicit images from a person’s inbox, users will only see a blurred image unless they elect to view the real deal.
Gallery Guardian takes censorship one step further by actually notifying parents of when their teens send or receive suspicious images via a social media app or text message. Founder and CEO, Daniel Skowronski, told Digital Trends, “We built a library of pornography to train our algorithm. To improve the detection, we also leveraged data sets that were mostly ‘selfie’ style images, since it’s the most common photo style taken with smartphones.” Though parents don’t get to view the offending content, the app does provide a head’s up so parents can intervene.
SafeToNet, a British startup, plans to use an algorithm and concept similar to Gallery Guardian’s. However, instead of alerting parents right away, the app gives teens several warnings before doing so. The warnings come in the form of a phone lock-out. After a fourth attempt to send or receive sexually explicit content — including text — the app will alert the parent.
As of right now, using AI in social media to filter and block dirty images and text is not a thing. However, it is promising to note that the technology is there. With a bit of a push, parents of future generations may have peace of mind that their teens are not and cannot engage in sexting. Until then, the only solution is good, old-fashioned investigating. Talk to your kids, check their phones, and make sure they know what is and isn’t appropriate behavior.
AI in Social Media Sources:
Digital Trends, Worried your kid is ‘sexting’? This AI-powered watchdog app can help
Gizmodo, This App Will Rat Out Teens Who Won’t Stop Sexting
OneZero, Medium, Can We Automate the Dick Pics Away?
Forbes: Tumblocalypse: Where Tumblr And Its Users Are Headed After The Ban