According to the nonprofit Digital Promise, AI literacy is rooted in three connected concepts: understand, use and evaluate. It posits that everyone in the educational system, from students to teachers to administrators, must understand what AI is and where it comes from, how to use it in their educational context and how to evaluate tools and use cases.
At the AI Show at the ASU+GSV Summit April 6, panelists from Digital Promise, Gwinnett County Public Schools in Georgia and the Kapor Center shared what this framework involves and how schools and districts have implemented AI literacy lessons to suit their needs.
UNDERSTAND
Shayla Cornick, director of research and development for ed tech at Digital Promise, said understanding the basics of AI is of foundational importance for being an informed user or procurer of the technology.
“Understanding AI is an essential component of AI literacy, because in order to make informed decisions about using AI and evaluating AI, you have to know what it is,” Cornick said. “You have to understand how artificial intelligence data sets are created and associations and automated connections.”
The best way for districts to add AI lessons, she said, is to embed them into strategic and instructional goals they already have.
“As I travel, every educational space looks different, it feels different, and I am always cognizant of time. It is a precious commodity,” Nicole Adell, associate director for school impact at Digital Promise, said. “I don’t want my district and school leaders feeling like, 'This is something else I’ve got to make some time for.'”
Adell said outreach to parents and caregivers to hear what they think about the role AI should play in education is an important part of this, as it also gives the community a chance to understand the basics around AI and what local schools are doing to keep data safe. One school she worked with recently in South Jersey held community outreach events every month for this purpose.
USE
Several panelists made the point that understanding AI is a prerequisite for using it. That way, students, teachers and leaders are equipped to make the first decision: to opt in or out. This could be for using AI at all, using a specific tool or using AI in a specific setting.
“We have to let people understand that it is OK to say no to the technology,” Shana White, director of computer science equity initiatives at the Kapor Center, said, qualifying that educators shouldn't feel like it’s a bad thing to opt out.
Introducing AI to reticent staff members tends to work best when it's done through existing meetings and systems, Adell said. For example, one school redesigned an upcoming staff meeting to incorporate discussing AI and trying out new tools.
Holloway said that teachers incorporating AI into lessons should ground it in real-world examples to encourage critical thinking that will be applicable even when students go off into jobs that haven’t been invented yet.
“If it’s siloed in just that content and just that one example, and it never leaves the classroom and never brings in real-world examples, then those students don't know how to then apply that learning,” she said.
Incorporating AI use into classroom lessons may involve discussions of what ethical use looks like. For White, this comes down to three questions: Does it harm you? Does it harm others? Does it harm the environment?
If it doesn’t pass any of the criteria on that test, she said, the question becomes: How can the user engage it responsibly?
In this way, learning about AI use ties into understanding how it works, as potential harms — such as adverse environmental impacts and exploitative labor practices — should be part of discussions about what AI is and where it comes from, White said.
EVALUATE
Digital Promise’s framework for AI literacy extends to the consideration of ethical personal use and the ability to evaluate AI tools on the whole. This largely includes knowing which questions to ask, either internally to oneself or out loud to a vendor, depending on one's role.
For students, parents and caregivers, White said relevant questions are: Who is creating this tool, and who isn’t? Who is benefiting from these tools, and who isn’t?
In addition to these questions, decision-makers in the education world, from classroom teachers to superintendents, should know additional questions to ask around data privacy, Holloway said. Vendors should be able to answer: Why did you create this tool? Is your tool appropriate for the age range in my building? Where does your training data come from, and what is my students’ data going into?
Gwinnett County Schools has an in-house team focused on procurement, and Holloway said they haven’t approved very many AI tools.
“We want our students, when it’s age appropriate, when it’s appropriate to the context, to have the experience, because application is important to understand how things work,” she said. “But we really want to do that in a protected environment.”
