The variety of districts that educated lecturers on generative synthetic intelligence greater than doubled from 2023 to 2024, in line with a brand new evaluation by the analysis and coverage agency RAND Corp.
Within the fall of 2024, 48 % of districts reported that they had educated lecturers on AI use—up from 23 % within the fall of 2023. RAND surveyed about 300 districts in 2024 and about 200 in 2023 and weighted their responses to make them consultant of all public college districts.
If districts comply with their plans for the present college yr, almost three-quarters can have offered AI coaching to educators by the autumn of 2025.
This skilled growth has been gradual to realize floor in colleges. Whereas generative AI turned mainstream in late 2022 with the launch of ChatGPT, 7 in 10 lecturers had not obtained any coaching on use AI within the classroom by spring 2024, in line with a nationally consultant survey from the EdWeek Analysis Middle. By fall 2024, although, 43 % of lecturers mentioned that they had obtained a minimum of one coaching session on AI, the EdWeek Analysis Middle discovered.
Each lecturers and principals are more and more utilizing AI of their day-to-day work. Whereas principals use AI instruments for administrative duties like making schedules, writing newsletters, or crunching massive datasets, lecturers are extra doubtless to make use of AI to generate lesson plans, customise classes for college kids who may study at completely different speeds and ranges, and even use it creatively—to emulate a historic or fictional character to show a selected lesson, for example.
Even so, most districts took a gradual strategy to introducing their lecturers to AI.
Along with fielding the surveys, RAND performed in-depth interviews with 14 district leaders on how they crafted their AI trainings in 2024. As an alternative of leaping straight into AI instruments and their use, district leaders centered first on tackling lecturers’ issues in regards to the know-how.
Of the 14 leaders interviewed, all however one mentioned they encountered “destructive views of AI and reported that some lecturers view it as a menace to conventional instructing strategies or a device for scholar dishonest,” in line with the report.
District leaders mentioned they offered coaching on the elemental workings of AI—and its limitations—to decrease lecturers’ anxiousness, and “shift away from an antagonistic, cheating-centered mentality” about AI.
Among the many 14 districts, trainings diversified in size and content material—from daylong trainings to “bite-sized” content material delivered by common emails and newsletters. All however one district made such coaching voluntary.
“I’m not attempting to push AI with my folks, however I’m attempting to present them the instruments in the event that they’re . I do know that as [AI adoption] begins occurring, it would develop organically and really shortly,” one district chief informed RAND researchers.
District leaders are inclined to craft their very own coaching
The district leaders interviewed by RAND most well-liked crafting their very own trainings, fairly than counting on exterior specialists, mentioned Melissa Diliberti, an affiliate coverage researcher at RAND and a lead writer on the report, in an e mail.
“This do-it-yourself strategy, although doubtlessly employed to fulfill particular district wants, additionally displays a shortage of exterior specialists who’re able to offering applicable coaching,” Diliberti mentioned.
Of the 14 district leaders interviewed, 11 took a DIY strategy to crafting their very own AI coaching. They pieced collectively completely different sorts of assets put out by tech firms like Google, in addition to training organizations like Digital Promise and the Worldwide Society for Know-how in Training.
Half of the interviewees mentioned they struggled to seek out coaching specialists who have been educated about utilizing AI in an academic context.
A time and useful resource crunch additionally pushed district leaders to drag collectively their very own coaching, within the absence of ready-to-use assets, the report said. Skilled growth in AI jostles for house amongst a number of different competing PD priorities, which might have an effect on the size and depth of such coaching, the district leaders mentioned.
District leaders additionally used their very own “destructive” experiences with AI trainings—the place they have been left overwhelmed—to streamline the coaching they provided so it addressed lecturers’ instant wants.
A suburban West Coast district, for example, first launched “play” classes to alleviate lecturers’ anxiousness about AI, earlier than regularly transferring onto extra instruction-based studying.
The “play” strategy, on this district and others, consisted of lecturers experimenting with an AI chatbot or platform and getting the hold of what AI can do. There have been extra centered classes, too, during which district leaders launched instruments like MagicSchool or ChatGPT to point out lecturers use them to plan classes or generate differentiated content material for college kids.
Chris Chism, the superintendent of the Pearl college district in Mississippi who was not related to the RAND report, additionally created his personal AI coaching module for lecturers. His coaching, too, emphasizes each experimenting with the know-how, and showcasing extra streamlined, direct makes use of for the classroom.
In an interview with Training Week final July, Chism mentioned he had created mini chatbots that ship important info to lecturers about particular subjects. One chatbot, for example, can generate lesson plans aligned with the state’s requirements.
Whereas district leaders’ trainings could also be extra in tune with a trainer’s wants, Diliberti mentioned the standard and what’s lined in these trainings may differ between districts.
There’s an fairness hole in AI coaching
Regardless of the general surge in AI coaching, the RAND report exhibits that such coaching is inconsistently distributed—in comparison with almost 67 % of low-poverty districts which have launched AI coaching for lecturers, solely 39 % of high-poverty districts have been in a position to do the identical.
RAND used districts’ plans to undertaking what the coaching hole may appear to be within the fall of 2025. Whereas almost all low-poverty district would have educated lecturers on AI use, solely 6 in 10 high-poverty districts would have executed so, if districts persist with their plans.
Faculties in low-poverty districts usually tend to have the time and assets to experiment with AI instruments and work out use them properly for educational follow. It’s not a luxurious that colleges in high-poverty districts could have, mentioned Diliberti.
This hole in coaching, or steering, on use AI persists among the many principal group, too, in line with a earlier RAND report revealed in February. 13 % of principals within the highest-poverty colleges mentioned they received steering, versus a few quarter of principals in lower-poverty colleges.
“Academics in higher-poverty colleges are much less doubtless to make use of AI, [as are] principals within the highest-poverty colleges. In order that in all probability drives whether or not the college system or the college itself is like, ‘Oh, we actually want to supply steering,’” Julia Kaufman, a senior coverage researcher at RAND and lead writer of the February report, informed Training Week on the time.
The quicker takeup of AI by lecturers in low-poverty districts signifies that lecturers and college students in high-poverty ones received’t have the identical alternative to study from AI instruments, the report says.
There are nonetheless no clear hyperlinks between better reliance on AI instruments and higher-quality instructing or studying, Diliberti mentioned.
“However to the diploma that we anticipate this is likely to be true, better uptake of AI in traditionally advantaged college settings does elevate issues that the rollout of AI into colleges may exacerbate longstanding fault traces of instructional inequality,” Diliberti added.
The report requires extra focused funding by federal and state training companies to high-poverty districts to acquire mannequin skilled growth in AI and domesticate connections to networks of specialists.
The report additionally recommended that districts with extra expertise in coaching lecturers on AI ought to share their experience with different districts by state and regional training networks.
The surge in coaching however, Diliberti mentioned it’s nonetheless too early to determine how precisely AI may match right into a district’s PD schedule for the yr.
“Proof from our district leaders counsel that their trainings are within the very early levels. [They are] extra centered on simply getting lecturers snug with AI instruments,” Diliberti mentioned. “Common trainer coaching on whether or not and greatest incorporate particular AI instruments into educational follow should still be a while away.”
