How to protect your kids from the dangers of AI - and how to actually use it for good

Originally Published on WFAA on 5/15/24

Numbers from Pew Research underscore the concerns. 95% of their teen respondents in a recent survey say they own or have access to a smartphone.

Countless new stories have sounded the alarm about the growing influence and risks of artificial intelligence – AI. The technology raises specific concerns for digital parents, concerned about increasingly lifelike deepfake images that can be used for cyberbullying and sextortion scams against their kids.

But like most technologies, it’s all about the application. Amidst complaints that big tech companies are sacrificing digital safety for profits, a slew of new firm say they’re working to make AI work for families, in ways that are simpler, safer and hopefully healthier.

“Our epiphany was, why don't we start kids that way right now?” asks Tim Estes, CEO of Angel AI. “And why don't we design the technology to be something parents own and control. So it's not beholden to any third party trying to influence their child.”

Numbers from Pew Research underscore the concerns. 95% of their teen respondents in a recent survey say they own or have access to a smartphone. But 76% of parents surveyed say it’s a priority to manage how much time their kids are spending on their devices.

There are increasing calls for parents to delay that access, with more studies indicating a relationship between device and social media usage and poor mental health outcomes for children. But experts like Estes argue there’s a middle ground. With an increasingly digital world, he says kids have to be prepared for it in ways that are age appropriate and give control back to their parents.

Estes cites a common critique – that for so many social media companies, the user is the product. Which means many tech companies are financially motivated to get as many users as possible online for as long as possible. Often, that means luring kids onto their platforms, with few incentives to create platforms with digital safeguards in place.

And that’s where AI comes in. The technology can allow for scalable technology to do things it would normally take an army of workers to do. For instance, Canopy is a parental control app using AI technology to identify harmful images and filter them, in real time, before they get sent to or from a child’s device.

“We're one of the few companies that have really trained an algorithm to identify explicit images and we have about nine patents,” says Canopy CMO Yaron Litwin. “I think for many parents, they understand the threats. They don't always understand that there's a solution. Our technology kind of comes in and is able to block that out, in real time before the child is exposed to that.

While Canopy uses AI to filer images, Estes describes the artificial intelligence powering Angel as a bridge for a family getting online. Their chat feature allows a kid to ask questions – about anything – with Angel offering up age-appropriate answers. Examples are wide-ranging, including why the sky is blue, how cars work, when’s the best time of day to exercise, etc. And if Angel encounters a question that’s beyond an expected age range, it’ll refer the child to a trusted grown-up. Parents can see their kids’ conversations with Angel, with the goal of allowing for oversight and better communication.

“We believe AI can be a bridge for that,” Estes says. “For parents and kids to understand each other better, understand the parent’s perspective on key issues, and understand the kid’s interests.”

Estes says like other forms of AI, Angel should get smarter over time, honing its interactions to be even more nuanced and helpful. And the goal isn’t to keep kids online for endless amounts of time – Angel will even gently nudge kids offline, so they’re not glued to their devices for hours on end.

“AI can be the intermediary that we've been missing for our children online. The important thing is the business model, the tech and its design, the product and its design. And ultimately the authority has to go back to the family.”

At a time with substantial focus on the risks of artificial intelligence, Canopy, Angel and other tools highlight the need for AI for good. The hope is that if they get this right, the rest of the industry will follow.

“We have to all come together to have a better generation than what we just gave the last generation, which is not healthy in the end,” Estes says. “This is about us playing our part in that.”