More school districts than ever are training educators on artificial intelligence (AI), according to a new RAND Corp. report. In 2024, 48% of districts provided AI training—up from just 23% in 2023. If current trends continue, nearly three-quarters of districts will have trained teachers on AI by fall 2025.
Despite this growth, many educators still lack formal AI education. A spring 2024 EdWeek Research Center survey found that 70% of teachers had received no AI training. By fall 2024, that number dropped to 43%, signaling progress—but also highlighting how much work remains.
Despite this growth, many educators still lack formal AI education. A spring 2024 EdWeek Research Center survey found that 70% of teachers had received no AI training. By fall 2024, that number dropped to 43%, signaling progress—but also highlighting how much work remains.
How Teachers and Principals Are Using AI
AI adoption is increasing in schools, but teachers and administrators use it differently:
However, many districts are taking a gradual approach to AI training. Instead of diving straight into tools, leaders first address teacher concerns—particularly fears that AI could replace traditional teaching or encourage cheating.
- Principals rely on AI for administrative tasks (scheduling, data analysis, newsletters).
- Teachers use AI for lesson planning, personalizing instruction, and even role-playing historical figures to engage students.
However, many districts are taking a gradual approach to AI training. Instead of diving straight into tools, leaders first address teacher concerns—particularly fears that AI could replace traditional teaching or encourage cheating.
DIY AI Training: Districts Craft Their Own Programs
Most districts aren’t waiting for outside experts—they’re building their own AI training programs. According to RAND researcher Melissa Diliberti, this reflects both a lack of available experts and the need for tailored solutions.
Of the 14 district leaders interviewed:
Chris Chism, a Mississippi superintendent, developed his own AI training, including mini chatbots that generate standards-aligned lesson plans.
Of the 14 district leaders interviewed:
- 11 created their own training, pulling resources from Google, Digital Promise, and ISTE.
- Half struggled to find AI experts who understood educational applications.
- Many used "play" sessions to help teachers experiment with AI before formal instruction.
Chris Chism, a Mississippi superintendent, developed his own AI training, including mini chatbots that generate standards-aligned lesson plans.
An Equity Gap in AI Training
While AI training is expanding, low-poverty districts are far ahead:
This gap extends to principals: A February 2025 RAND report found that only 13% of principals in high-poverty schools received AI guidance, versus 25% in wealthier districts.
Julia Kaufman, a RAND researcher, noted:
“Teachers in higher-poverty schools are less likely to use AI, [as are] principals in the highest-poverty schools. So that probably drives whether the school system… provides guidance.” If this trend continues, students in high-poverty schools may miss out on AI’s potential benefits.
- 67% of low-poverty districts offered AI training in 2024, compared to only 39% of high-poverty districts.
- By 2025, nearly all low-poverty districts may provide training, while only 60% of high-poverty districts will.
This gap extends to principals: A February 2025 RAND report found that only 13% of principals in high-poverty schools received AI guidance, versus 25% in wealthier districts.
Julia Kaufman, a RAND researcher, noted:
“Teachers in higher-poverty schools are less likely to use AI, [as are] principals in the highest-poverty schools. So that probably drives whether the school system… provides guidance.” If this trend continues, students in high-poverty schools may miss out on AI’s potential benefits.
There’s no proof yet that AI improves learning outcomes—but if it does, the current rollout could deepen existing inequities. Diliberti warns: “Greater uptake of AI in historically advantaged school settings raises concerns that AI could exacerbate longstanding fault lines of educational inequality.”
To close the gap, the report recommends:
Despite progress, most AI training remains introductory, focusing on comfort rather than mastery. As Diliberti notes: “Regular teacher training on whether and how to best incorporate AI into instruction may still be some time away.” For now, schools must balance innovation with equity—ensuring all students and teachers benefit from AI’s potential.
To close the gap, the report recommends:
- Targeted federal/state funding for high-poverty districts.
- Sharing expertise between districts via state networks.
Despite progress, most AI training remains introductory, focusing on comfort rather than mastery. As Diliberti notes: “Regular teacher training on whether and how to best incorporate AI into instruction may still be some time away.” For now, schools must balance innovation with equity—ensuring all students and teachers benefit from AI’s potential.