By the students, for the students: Sierra High School represents Colorado in developing national AI framework
To many, the recent, rapid development of generative artificial intelligence and its lack of frameworks and oversight may evoke comparisons to the unrestrained Wild West in the 1800s or the economic potential of the dot-com boom in the early 2000s.
Recognizing AI’s role in the future, students from all 50 states convened over the summer on a recreation of the U.S. Senate chamber at the Edward M. Kennedy Institute in Boston to develop a framework for the technology’s responsible use in public education.
This student senate featured Marissa Donjuan and Logan Arviso from Harrison District 2’s Sierra High School as Colorado’s representatives.
“I noticed that students have very differing views on AI,” Donjuan said. “There are some students who wouldn’t use it for anything. There were some students who, like me, use it for things like brainstorming and there are also students that, like, rely on AI and use it heavily.
“So, there’s a lot of different ways that students were using AI that I think helped the policy be broad enough to fit in everyone’s perspective.”
Deliberations included which grade to introduce AI to students, how to educate them about ethical use, what appropriate punishment should be for unethical use and how teachers should use it.
Following two days of subcommittee sessions, amendments and compromises, the STUDENTS FIRST Act passed on an 82-16 vote on July 19 and will be shared with superintendents and school leaders nationwide.
Rather than banning AI in schools, the students established boundaries for it to support education while prioritizing original student work, professional judgment and human connection.
The proposal would require AI-literacy instruction when students begin using digital devices in school; permit using AI for brainstorming, studying and editing assignments in high school; and give students the right to appeal accusations of improper AI use. It would prohibit AI use to complete artistic or written assignments and generate materials to bully or falsely accuse individuals.
The act also preserves the right of high schoolers and educators to refrain from using AI.
The convention was part of America’s Youth AI Festival, a three-day retreat hosted by the nonprofits MIT RAISE, Day of AI, the Edward M. Kennedy Institute and AASA, The School Superintendents Association.
D-2 Superintendent Wendy Birhanzel, an AASA member, selected Donjuan and Arviso to be Colorado’s representatives. Their principal, Connor Beaudoin, also attended the festival and discussed developing policies and curriculum regarding AI at the administrative levels.
“States were able to share their journeys with us,” he said. “Some folks were still in their infancy phase and others were fully diving into this at the district level (and) in the classroom.”
The festival and partnerships came together after educators recognized the need to improve AI literacy and encourage civic engagement. As students crafted their legislation, educators learned about effective AI models in countries like Colombia and Rwanda.
Last year, D-2 adopted policies on AI use and maintaining academic honesty.
Beaudoin said the new policies support the district’s goal of developing students’ transferable skills, since they will likely use these emerging tools following graduation.
“Our district does not want to shy away from AI. We want to walk alongside it because it’s evolving so rapidly,” he said.
Day of AI’s executive director, Jeff Riley, is also a veteran educator with over 30 years of experience working in Massachusetts and Maryland.
“We missed the boat on cellphones and the internet years ago,” he said of their impacts on public education. “I don’t want to do this again.”
Despite concerns over academic honesty, Riley said AI has already yielded benefits in the education field. He noted that teachers have used it for administrative paperwork, drafting emails to students’ parents and other minutiae, freeing up their time to focus on their work in the classroom.
He said AI can be used to differentiate and personalize instruction for each student. For a class of students with various reading proficiency, instructions and summaries could be generated for each student’s reading level rather than unintentionally lowering expectations with a one-size-fits-all approach, which he says has occurred in the past.
With AI now commonplace in most people’s lives, Riley said the biggest focus going forward is AI literacy. He, for instance, uses AI daily and argues with the popular ChatGPT “about three times a day” over inaccurate prompts.
He said he’s able to do this because of his own knowledge base that younger students don’t have. This, he said, presents an opportunity to develop their critical thinking by remaining skeptical of AI-generated responses and verifying information with multiple AI programs and sources.
Riley added that all students now are “digital natives,” which refers to those who have access to and use digital technology at an early age.
“I think we should start by asking them what they think,” he said before any federal framework develops. “That can evoke a lot of discussion.”
Arviso and Donjuan, now entering their senior year at Sierra, intend to continue educating themselves and others about AI as it continues to develop. Donjuan, who sits on the D-2 Board of Education as a student representative, plans to present the framework to board members to refine the district’s practices.
Both admitted that the adopted framework — like many bills passed in the U.S. Senate — isn’t perfect but that it represents an important first step.
“I feel like the main reason why we voted ‘yes’ on it was that it’s the beginning of something,” Arviso said. “This is a building block for a bigger and better bill.”