High school students will soon have new opportunities to build and analyze artificial intelligence technologies. CodeAI and OpenAI announced a year-long partnership to bring technical training, student challenges, and teacher support into classrooms. The initiative aims to shift the educational focus from simply using AI tools to understanding the underlying systems.
What Happened
The partnership focuses on programs that explain how artificial intelligence works. According to the announcement published on the EdTech Innovation Hub, OpenAI will support CodeAI’s Hour of AI initiative and work as a resource for its year-long AI Foundations course. OpenAI employees will also mentor students in CodeAI's new Builders Challenge, where high schoolers design their own AI projects.
The two organizations are also establishing a joint advisory council of specialists in child development, learning science, and youth public policy. This council will work to promote responsible AI use and identify classroom risks. As we previously reported, safety remains a major concern for parents, especially as OpenAI expands its defensive presence in schools to counter new digital threats.
CodeAI’s leadership describes this collaboration as a "glass box" approach. Instead of treating AI as a black box that cannot be inspected, the program teaches students to look inside the technology to find bias and recognize limitations.
The Bigger Picture
This partnership addresses a major gap in secondary education. Research from CodeAI indicates that only 16% of high school leaders believe their students are currently learning the technical concepts required to understand AI. CodeAI also reports that 75% of high school students believe AI literacy will be important for their future careers.
However, broader research reveals that older students are skeptical about these technologies. Technical skills are increasingly sought after. According to Handshake employment data, non-computer science majors in fields like business and biology are actively building AI skills, yet many students remain highly skeptical.
A 2026 survey published by the University at Buffalo Career Design Studio found that nearly one-third of graduating college seniors believe AI skills will have "little or no importance" to their future careers, and half are not building them at all. Research featured in Inside Higher Ed reveals that 40% of students worry about becoming overly dependent on AI, and 25% actively resist using the technology due to ethical and environmental concerns.
To bridge this gap, educators are looking toward structured frameworks. The finalized AI Literacy Framework for Primary and Secondary Education, developed by the OECD and the European Commission, provides schools with classroom-ready guides. This interactive framework shows that literacy requires students to avoid anthropomorphizing AI, meaning they must learn that these tools do not "think" or "understand," but operate entirely by analyzing patterns in data.
What This Means for Families
For parents and educators, teaching AI now goes beyond writing prompts. The focus has moved to "algorithm auditing," which allows students to detect systematic biases.
According to researchers at the University of Pennsylvania Graduate School of Education, students do not need complex programming skills to analyze AI fairness. By testing different prompts and evaluating the outputs, students can use basic statistics to uncover algorithmic biases. This method helps dismantle cultural stereotypes embedded in models that often rely on narrow training demographics, as detailed by research on teaching bias as an equity practice.
What You Can Do
To support this learning at home, parents can encourage children to audit the generative AI tools they use. For example, children can test different prompts to see how a system's responses change, looking for patterns, omissions, or biased assumptions in the generated text and images.
Parents and teachers also do not need expensive software to teach these concepts. Educators can use free, "unplugged" lesson plans from the AILit Framework to explain data patterns and algorithms without a screen.
Finally, parents can help students avoid anthropomorphizing technology by encouraging them to use precise language. Children should be reminded that a chatbot does not have feelings or knowledge; it simply predicts the next most likely word based on its training data.