https://www.commonsense.org/education/articles/chatgpt-and-beyond-how-to-handle-ai-in-schools

As technology evolves and becomes more sophisticated, it's understandable that we're uneasy about actual and anticipated challenges. Here are some current concerns:

• Abuses 

  • Deepfakes are pictures, videos, and audio files that look or sound like someone you know (a celebrity, a political figure, or even a family member). This can also include nude images or pornographic videos generated with AI and without a person's consent. Kids are already using so-called "nudifying" apps to generate nude images of classmates.

How to Handle Abuses

• Get in front of potential problems by teaching AI literacy and having discussions about the technology's harmful applications.

• Make students aware that deepfakes exist online already, so it's good to be skeptical about audio and video that's especially shocking or shared widely to get an emotional response. People can still spot some imperfections, but they're getting more difficult to identify!

• Talk about consent when it comes to content. Just like you should ask before taking a picture or video of someone, it's important to get someone's approval before even thinking about feeding their image into an AI tool. Encourage kids' empathy by asking how they might feel if someone posted something embarrassing about—or of—them. 

• Be clear about the consequences of using AI to generate pornographic images. Tell students that they may see apps to create nude pictures advertised on platforms like TikTok. Though they may be curious or think it's funny (because the pictures aren't "real"), using AI to generate nude pictures of someone is harassment and likely illegal. It doesn't just harm the victim—law enforcement could get involved. Victims should tell a trusted adult, report to authorities, and can also report the incident to CyberTipline.org.

• Biases and misinformation

  • AI can only learn from its source(s), so it takes on the biases, misinformation, and problematic content of the original material. And if the team of developers isn't representative, it's almost guaranteed that implicit bias will be woven into the framework of the tool, as facial recognition has illustrated.

How to Handle Biases and Misinformation

• Show students examples of images generated by AI tools with specific prompts, and discuss the biases illustrated in the output.

• Use media (like sci-fi movies, shows, and books) to introduce discussions about algorithmic bias, the flaws of facial recognition, and more. 

• Remind students to include requests for source links when they're using AI to do research. (Some tools will supply links to their source material, when requested.) Ask students to check each link.

• Tell students about AI hallucinations. Explain that AI can combine information from its data sets to generate complete fabrications. AI tools can also claim to feel love, experience various human senses, and more. So it's important for students (especially younger ones) to understand that an AI chatbot may sound like a real person, but it's designed that way—by real people.

• Environmental impacts 

  • Generative AI requires an enormous amount of energy and other resources, including fresh water as a cooling mechanism. In an age when we already have many climate concerns, the growing use of AI only adds to these issues.

How to Handle Environmental Impacts

• Give students age-appropriate context about global climate change and Earth's resources. 

• Inform students about the heavy resource load required to use generative AI. How about cellphones? Computers? How does advanced technology impact the planet?

• Highlight the people working to limit these negative impacts and find solutions to environmental issues.

• Ethics 

  • Because AI tools scrape content from a wide variety of sources, the material produced is a mixture of many other people's work, and there's often no consistent or complete credit for creators. Plus, data privacy is a murky and multi-layered issue when it comes to generative AI.

How to Handle the Ethics Involved

• Talk about the sketchy ethics of how generative AI works, but also get into a larger discussion about the impacts of technological developments and how we could/should be handling them.

• Talk about the existing digital divide and what new developments will mean for people who can't afford them. What happens as tools like this start to cost money for more features?

• Discuss the protection of intellectual property. What guardrails do we need if we want to get ahead of potential problems?

• Talk about data privacy. What information does AI currently collect, and what might it gather in the future? How does it use the information you give it? Talk about the potential impacts of an AI that can piece together data about you from everywhere online.

• Plagiarism

  • One of the biggest worries for teachers is plagiarism. Already, students are handing in AI-generated essays as their own. 

How to Handle Plagiarism

• Play with AI platforms yourself to understand what they can and can't do. Plug in your prompts and test them.

• Get to know students' writing as much as possible.

• Do some writing in class, staying mindful of some students' limitations to do so.

• Use formative assessment to get snapshots of progress over time.

• Evaluate your prompts and try to include a student-centered approach that features emotional intelligence and experience. As one teacher says, "If you don't want generic answers, don't ask generic questions."

• Test AI tools together and discuss their limitations.

• Include a "trojan horse" word or phrase in your assignment that isn't visible to the student—but you can use this keyword later to see  if the student pasted the prompt into an AI tool.

• As a last resort, there are tools you can use to detect AI, but be aware that they aren't perfect and flag plagiarism where it doesn't exist almost half of the time.