Artificial intelligence continues to present a growing concern for our children and our families.  It’s not going away, and it seems to become harder and harder to detect (maybe that reveals my age). And, as you know, it can be misused to spread dangerous and false information. Even more, artificial intelligence can present misinformation in a very convincing manner, even with pictures that look realistic.

Still, many people go straight to AI for their information. They simply ask some chatbot to search the web and provide the answer, then trust that answer as accurate. But it may not be accurate. We witnessed a vivid example of this when Grok “debunked 20 of the president’s false claims.” But after Musk updated the program, it “began spewing out virulently antisemitic tropes…and celebrating political violence against fellow Americans.” Live Science Plus has identified 32 times artificial intelligence “got it catastrophically wrong.” For instance, one chatbot encouraged business owners to perform illegal activities such as stealing workers tips or paying less than minimum wage.  Another chatbot was giving harmful advice for those with eating disorders. An Amazon AI recruiting tool was found to discriminate against women. The list goes on.

Potentially misleading AI information has even made it into our government. USA Today quoted “Ivan Oransky, co-founder of Retraction Watch,” who said one MAHA report seemed to share characteristics similar to other AI-generated work. AI papers “tend to hallucinate references,” he said. “They come up with references that share a lot of words and authors and even journals, journal names, but they’re not real.” Unfortunately, our government may use such documents for policy decisions if we don’t carefully discriminate potentially false information provided by AI.

AI does have positive uses. But, as you can see, AI also has the potential to mislead those who do not think critically about what they read on social media and other web search engines. We need to protect our families from this threat, especially our children.

One thing you can do to teach your children the potential for AI misinformation or misuse by others is to engage AI Puzzlers with them. AI Puzzlers was developed by the University of Washington to show children an area where AI systems “typically and blatantly fail.” In the AI Puzzlers game, users have the opportunity to solve simple visual reasoning puzzles. After the user (like you or your child) know the answer, they can ask various AI chatbots to complete the puzzle. Researchers discover that the chatbot often fails at the puzzle, even though the children easily completed them. When AI did get the answer correct, it often gave an inaccurate explanation for how it found the right answer. In other words, it gave confident misinformation. Even when it failed, AI offered a confident explanation for why its choice “is correct.”

By engaging in AI Puzzlers with your children, you can begin to teach them to question information obtained through chatbots, AI, and even social media sites that might utilize AI generated information. Of course, this isn’t a perfect solution. It’s a beginning to teach our children critical thinking when it comes to screen time. Other important ways you can help include:

• Monitor your children’s screen time, including their social media use and web searches. I know this is unpopular with some families. However, parents have a responsibility to manage their children’s safety. You can better assure their online safety if you have access to their online materials. This will involve sharing passwords and checking their accounts.
• Set boundaries and limits on screen time. You may choose to have times when no one has their phone or other screens available to them. For example, a no-phones-at-the-table-when-eating-together rule makes mealtimes a web-free and phone-free space. You may also designate sleep time as a web-free and phone-free time. Younger children may charge their phones in a common area overnight. No other screens in the bedroom.
• Spend time discussing current events with your children and researching various sources of information about those events. Teach them how to seek out accurate information.

Of course, each of these suggestions will change as your child ages, matures, and grows in their ability to manage their screen time and identify truth in a more independent manner. But they provide a starting point.

Since we are all learning about managing AI as it grows, I wonder…what have you found helpful in teaching your children to think critically about information they find on social media, the web, and AI?