The most useful education research this month is not about a new app. On October 14, the Walton Family Foundation, Gallup and The Mattering Institute released a survey of 1,952 U.S. K-12 educators, conducted in late July and August 2026. Its headline finding is simple: 59% of educators used AI tools during the 2025-26 school year, for preparing lessons, building materials and assessments, grading, feedback and administrative work. AI in the staffroom is no longer an experiment. It is ordinary.
The encouraging half of the findings is about time. Among educators who used AI, 64% said it freed up time for the parts of the job that matter most to them — working directly with students. The detail matters: 13% said it gave them a lot more time, 51% somewhat more, 29% felt no effect, and 7% said the tools actually cost them time. Read honestly, this is neither the revolution of the brochures nor a failure. It is a real, uneven gain — the kind that depends entirely on what happens next.
The other half of the research is the warning. The same survey programme measured what its authors call mattering — whether educators feel valued, and able to add value. Barely a third of educators, 35%, said they matter a great deal at their school. One in five said they do not matter much or at all. And among educators who lack that sense of mattering, 68% reported being frequently burned out, and fewer than half planned to remain in education. A chatbot can return an hour to a teacher's week. It cannot return the feeling that the week was worth spending.
The survey is equally clear about where that feeling comes from. Eighty-five percent of educators said that helping a student through something difficult makes them feel they matter a great deal; 74% said the same about seeing students perform well. Mattering flows from students — not from supervisors, not from software. Which reframes the whole AI question for school leaders. The value of the technology is not the hours it saves. It is whether those hours travel back to students, or get quietly refilled with fresh paperwork, new reporting formats and additional duties invented to fill the gap.
There is a second gap the research exposes: guidance. In Gallup's companion study published in May 2026, only 18% of teachers reported receiving any formal guidance from their administrators on how AI should be used at work, and across common tasks roughly a third received no guidance at all. An international Delphi study published this month in Computers and Education Open — 70 experts from 18 countries, led by Kumamoto University — reached the matching conclusion from the other direction: the top priority for AI in education is not more tools, but AI competencies for learners and educators. The tools have arrived ahead of the training, and teachers are paying for it in improvisation and anxiety.
None of these numbers are Pakistani, and we will not pretend they transfer directly. The classrooms where we run AI-awareness sessions — low-resource schools in Karachi, and outreach in rural Khyber Pakhtunkhwa — differ from American schools in class size, in shared and scarce devices, in electricity and connectivity, and in language: most AI tools still think in English first, while most of our classrooms think in Urdu and regional languages. But the pattern underneath the percentages is one we recognise from our own sessions. A teacher buried in registers, exam clerical work and duplicated reporting does not need a lecture about the future. She needs an hour back — and the training and permission to spend it on the child who is falling behind.
For school leaders anywhere, the practical agenda writes itself from the findings. Decide in advance what saved time is for, and protect that decision when new demands arrive. Involve teachers in choosing the tools — teachers use what respects their judgment, and quietly abandon what does not. Provide real, formal guidance rather than logins and good wishes. And measure the right outcome: not hours saved on a spreadsheet, but whether student contact, feedback quality and teacher morale actually moved. A school that fills every AI-saved hour with more administration will get the efficiency. It will also keep the burnout.
Our own work starts from the teacher's week rather than the technology's promise — training first, tools second — because that is what the evidence, and our classrooms, keep teaching us. The deepest lesson of this month's research is not really about artificial intelligence. It is that a teacher's sense of mattering is built by students and by school culture, and no vendor sells either. AI can clear the desk. Only people can make the person behind it feel the work is worth doing. Schools that understand both halves of that sentence will keep their teachers. The rest will automate the paperwork and wonder why their staff still leave.
Sources: Walton Family Foundation, Gallup and The Mattering Institute, 'Teachers Using AI Say It Frees Up Time for What Matters Most' and 'The Mattering Gap Among K12 Educators', released 14 October 2026 (news.gallup.com; waltonfamilyfoundation.org). Gallup, 'Most Teachers Receive No Formal Guidance on AI Use', May 2026. Majumdar, R. et al., international Delphi study on AI integration in education, Computers and Education Open, October 2026 (reported by phys.org, 14 October 2026). All survey figures are as published by Gallup; full survey methodology is available in the original releases.
