Common Sense media weighs in why AI isn't kid safe

After seven right now, Google threatening to cut off access to its search engine for schools that do not opt in to the company's generative AI tools. In a blog post, a tech giant announced plans to automatically shut off Google search access to students if a school system opts out or restricts its AI tools. Critics say the policy forces schools and young students into accepting AI integration and data collection under the threat of losing basic Internet search. Google says the changes gives schools more control over how students use AI. The change goes into effect next year to help us better understand the potential outcome to students and educators. We're joined by Jim Steyer, CEO and founder of Common Sense Media. Thank you so much for joining us, Jim. Great to be here, Andre. All right, so your institute rated ChatGPT for teens an unacceptable risk and called for an under-18 pause. What is your advice to superintendents facing this ultimatum when teens' AI protections are visibly falling apart here? Say no to Google. I mean, it's that simple. This is really pretty. Outrageous behavior by Google basically search is the front door to the Internet. We know that that's true for students and it's also true for all of us. And Google shouldn't hold search hostage to push their AI platform that we rated unacceptable for kids as necessary in the classroom. So. This is a Faustian bargain. This is not a good choice for schools, and Google needs to change its policy, period. Jim, both Google and OpenAI claim their classroom tools promote learning rather than cheating. Yet your unimproved ChatGPT's study protections can be easily bypassed. Can you tell us about that? Sure. We run the only youth AI safety institute in the world at Common Sense Media, and we do rigorous product testing of things like ChatGPT and, quite frankly, Google Search. And what we found on the Google Search risk assessment about a month ago is that 100% of the time, the Google app actually did kids' homework for them. So you just plug it in, and instead of It does the work for the students. So that's just unacceptable. And Google needs to fix its problems. The idea that they would then tell school systems you can't use Google products unless you use our unacceptable AI application is crazy. And I think that schools will say no. I think superintendents here in the Bay Area and elsewhere will say no. And the bottom line is. Google has to change, not schools. It's about time they serve the students properly. You know, Google says Gemini is actually an educational aid and a tutor here. Meantime, the search and AI are technically inseparable, according to Google's core explanation of this here. But what are your thoughts here? Because on this issue, your testing revealed that ChatGPT missed more than one in four crisis referrals for self-harm and disorder eating here, sending zero alerts to parents on new accounts. So this is potentially an issue here when the companies say, hey, our system will be tracking these type of things. No kidding. This is an issue, Andre. This is an issue for every parent, every school, every every educator. Look. Some of these big AI companies are in such a land rush in the arms race to win the AI battle that they're just putting out products that aren't safe yet. That is definitely true for Google search. And as you mentioned, we found big problems with the latest version of ChatGPT for teens. So the truth is the companies need to be held accountable. They need to get their act together before they push these products on kids in schools. It's that simple. And Hopefully, legislators here in California and elsewhere will hold them responsible for this. But in the meantime, common sense is going to prevail on this. Google will change their policy if they get enough pressure from schools and superintendents and, frankly, from the state of California. And kids will be the winners if that happens. Right now, Google is not behaving responsibly, period. You know, OpenAI promised behavioral age estimation, but what did your organization find when it looked into that claim? It found that they didn't do it very well yet. These products are still in development. Just think of how recent the phenomenon of AI is. And the fact that they're just rolling out product after product and new version after new version without appropriate safeguards is a big problem. Look, Common Sense is the biggest group in the United States focused on these issues. We're not anti-tech. We believe tech and AI used appropriately can be great for kids and families in school. But right now, a number of the products have significant flaws, significant risks for kids. And you just can't push these products that have all these flaws and risks on young people without fixing them. And that's what we're calling on Google, OpenAI. And the entire AI industry to do, you know, what Google says here, they explain that their strategy is evolving here and their tools allow students to to to work through learning challenges step by step. But, you know, your organization is looking into it from from an outside perspective here with a more critical eye, I suppose, here. But, Jim, thank you so much for joining us this morning. Jim Steyer, CEO and founder of Common Sense Media. We thank you so much for joining us this morning. Good to see you, Andre. Pleasure.