Salamanca City Central School District, a rural Western New York district, paused a $57,590 pilot with Realbotix that would have placed a humanoid robot named Sally in a high school classroom this fall, New York Focus's Melissa Manno reported July 24. The robot, designed to teach coding, robotics, and artificial intelligence courses, came paired with a laptop-accessible AI avatar that could give homework feedback and translate into more than 100 languages. The district announced the pause hours after state Education Commissioner Betty Rosa sent a letter warning "the Department remains concerned about the effect that a humanoid robot may have within a high school classroom." New York State United Teachers President Melinda Person went further, saying "a robot built by a company associated with sex dolls has no business in our classrooms," a reference to Realbotix's 2024 acquisition of RealDoll's parent company.
Nobody in this fight disputed the coding, robotics, or AI curriculum Sally was built to teach. The objection was entirely about data privacy, optics, and who built the thing, not what it was going to teach your students. That is the actual lesson for any science or STEM teacher fielding a vendor pitch this fall: the curriculum content is rarely what gets a pilot paused. The company's data practices and ownership history are.
The Wakulla County School Board in Crawfordville, Florida approved an AI use policy at its Monday meeting ahead of the upcoming school year, WCTV's Sedrick Bolton III reported July 28. The guidelines allow AI use districtwide but bar uploading personally identifiable student information into any AI tool without district authorization, require employees to review AI output for accuracy and bias rather than rely on it outright, and prohibit students from using AI to complete assignments in place of their own thinking. Superintendent Richard Myhre framed the policy as incremental: "The idea behind AI is to use it to strengthen our processes. It's not going to replace creativity. But there are some instances where we think its use would be an enhancement on our efficiencies." The district says it will amend the policy as state legislation on AI develops.
Wakulla County approved this policy the same week Salamanca's robot pilot became a statewide controversy in New York, and it generated zero public pushback. The difference is what each policy actually governs: Salamanca endorsed a specific product, a physical robot from a specific vendor. Wakulla wrote rules for behavior, what any tool may and may not be used for, regardless of which product a teacher picks up next. A behavior-first policy travels with the technology as it changes. A product endorsement doesn't survive contact with the first privacy question.
Austin Sarat, the William Nelson Cromwell Professor of Jurisprudence and Political Science at Amherst College, wrote in The Conversation on July 24 that college professors' now-routine complaint about students who can't finish a book has a documented cause below the college level. Sarat cites a 2024-2025 survey of 1,250 high school English teachers finding they assign an average of 2.81 books a year for students to read start to finish, down from an average of 4 books in a 2008-2009 survey of 400 English teachers. Some high school students are assigned no complete books at all, only excerpts, poetry, and news articles. Sarat also cites National Endowment for the Arts data showing the share of adults who read at least one book in the past year fell from roughly 55% to 49% over the past decade.
If your unit plans lean on excerpts because that is what fits the pacing guide, the district's own assessment calendar, or what students seem able to sustain, the survey data says you are the norm, not the exception. The norm now has a documented downstream cost: professors at highly selective colleges report students arriving unable to sustain a long, complex text, a stamina that no single excerpt is designed to build. The shift didn't happen in college. It happened in high school English classrooms, one unit-planning decision at a time.
A study published July 27 in JAMA Pediatrics, led by Y. Tony Yang of George Washington University, analyzed cellphone policies across all 50 states and Washington, D.C., and found 42 jurisdictions now have statewide laws or policies restricting cellphone use in K-12 public schools, up from a single state, Florida, in 2023. Of those 42, 32 (76%) enacted a ban during instructional time, including 15 bell-to-bell bans, while school-level restrictions actually declined between 2009-10, when 91% of public schools prohibited cellphone use, and 2021-22, when only 76% did. The study calls the statewide shift "a major shift in educational governance" and notes it is bipartisan, but states plainly that direct evidence linking cellphone use to declining youth mental health remains difficult to establish, even as it confirms cellphone-related task switching impairs academic performance. The nine jurisdictions without a statewide policy are Montana, Wyoming, South Dakota, Illinois, Mississippi, Pennsylvania, Delaware, Massachusetts, and Washington, D.C.
If your state hasn't mandated a cellphone policy, your building is likely deciding this on its own, and the research base available to justify a ban is stronger on distraction and academic performance than on the mental health case that usually leads the public argument. Utah Governor Spencer Cox called a bell-to-bell ban "the single most important thing that we can do for our students in our schools today," but the study's own authors say that specific claim runs ahead of what the evidence currently supports.