PROVIDENCE — Kevin Burke knew his social studies students were using AI. The question was what he should do about it.
“If I found work to be suspicious, I tried to put it in multiple platforms” that check for AI, said Burke, who taught high school social studies in Newport until earlier this year. “The percentages sometimes varied a lot. You can’t really prove it.”
Burke, who will switch to teaching English as a Second Language in Westerly later this month, said he’s open to using artificial intelligence in the classroom, but has so far seen little guidance in terms of best practices.
“Right now, schools are just behind the ball and trying to catch up, because it’s moving quicker than people can really keep up with it,” said Burke. “It’s coming full throttle, and it’s really the responsibility of educators to teach students how to use it responsibly, ethically, and for their betterment without misusing it.”
The R.I. Department of Education late Friday put out its first set of guidance on AI use in schools, giving public schools the clearest green-light yet that they should start exploring artificial intelligence, while also detailing pitfalls including inaccurate information, perpetuation of bias, and academic dishonesty.
The 53-page document encourages districts to start teaching AI literacy, vetting, and incorporating AI tools in the classroom, using AI to streamline tasks, and setting up official policies to use the technology.
“Artificial intelligence is not the future for our schools — it’s the present, and our goal is to ensure it enhances teaching and learning to unlock our students’ full potential,” Education Commissioner Angélica Infante-Green said in a statement.
School starts in some Rhode Island communities next week, while others return after Labor Day.
A survey conducted by RIDE found 20 percent of students say they are already using AI for schoolwork, while 6 percent of educators said they were using it at work.
School districts don’t have to follow the guidance, and can still create their own policies — including banning AI — but now have a roadmap for potential policies after spending several years with little direction.
More than half of states have now adopted some sort of AI guidelines for schools, according to AI for Education, a group that advocates for generative AI use in schools.
Massachusetts launched an effort earlier this year to implement AI in schools through 2027, and sent a letter to districts on Monday with guidance documents for how to create a policy. The Massachusetts Department of Elementary and Secondary Education left the decision of “when, how, and why” to use AI up to the individual districts.
Connecticut also issued guidance earlier this year, but said it could be “inefficient and burdensome” for each school district to create and maintain its own policies.
“District administrators and school board members should consider whether AI truly ‘breaks’ existing policy,” Connecticut’s guidance says. “Ways of addressing plagiarism and cheating, harassment, bullying, and data privacy should remain consistent, even if new technologies may introduce new challenges of understanding and application.”
Narragansett Superintendent Peter Cummings said he was waiting for the RIDE guidance before creating a policy in his district, and will come up with one this fall now that it’s out.
“I think it gives some solid, big ideas of what should be included,” Cummings said. “Students are using it outside of school. We want to introduce them to using it intelligently and safely as part of their education and digital literacy.”
Some Narragansett teachers are testing a tool called Magic School AI for lesson development, after Cummings said the company signed a student data privacy agreement with the district. Open programs such as ChatGPT are banned from the district’s network, out of privacy concerns.
The new guidance acknowledges the privacy issues, and urges school districts to verify with AI vendors that they are complying with privacy laws and have clear policies about data retention.
Districts are not required to have an AI policy, RIDE spokesperson Victor Morente said.
Some teachers remain unconvinced that AI should be implemented in school, and 78 percent of educators surveyed by RIDE said they were concerned about students using the technology to cheat.
“Students are using it to get out of engaging in any critical thought or hard work or higher-level thinking,” said Katie Behan, a history teacher who is starting a new job at New Bedford High School in Massachusetts this fall.
She previously taught in Newport, and said she initially tried to teach students to use AI responsibly. But she eventually abandoned the effort, finding students were not learning how to properly research or write.
“I’ve given up on trying to help students learn how to use AI,” Behan said. “They’re not tools, they’re crutches.”
She said despite showing students the pitfalls of AI, including inaccurate information generated by the chatbots, they continue to depend on it to write papers that are turned in with poorly-sourced information.
“It’s very tantalizing to just put the prompt into an AI chat and just have an essay formulated,” she said. “You’re going to finish high school not being able to write a simple 5-paragraph essay.”
She tells her students: “Be very wary of anything you are being given for free,” noting that the generative AI programs are collecting data on them and may not always be free to use.
The RIDE guidance does not suggest an age limit for using AI, but notes that there are “developmentally appropriate” uses for different age groups. Kindergarten through 2nd-grade students should use it only under the direct supervision of a teacher, and should be taught that AI is not a real person, the guidance says.
High school students could use it more widely, the guidance says, such as for polishing papers, brainstorming ideas, research, and more.
“We don’t want students in elementary school doing a lot with AI, because they’re still mastering the fundamentals,” Cummings said. “We don’t want AI to get in the way of that.”
To combat cheating with AI, Burke said he started emphasizing the entire writing process, not just the final paper, and had students do work in class on paper and not just on computers.
And while he would like more tools from RIDE to combat plagiarism, he doesn’t want to ban the technology from his classroom.
“Banning it is just holding back the future,” Burke said. “This is coming full-bore.”
Steph Machado can be reached at steph.machado@globe.com. Follow her @StephMachado.