Big tech pitches AI for teacher burnout. Educators call it a Band-Aid
From frequent burnout, increased workload and student behavioral problems, America asks a lot of its teachers. It’s little surprise the nation is grappling with a massive educational labor crisis. More than 400,000 teaching positions across the country are either vacant or filled with underqualified staff, according to data from the Learning Policy Institute.
Martina Meijer, an elementary school teacher from Brooklyn, New York, has witnessed the strain firsthand. Over her 18-year career, she has watched class sizes swell beyond state limits, administrative burdens skyrocket and underfunded schools forced to misassign teachers to subjects outside their training.
Artificial intelligence giant Anthropic believes it has a tool to ease that burden. The company recently unveiled Claude for Teachers, a specialized initiative designed to offset the heavy administrative load teachers carry. In its announcement, Anthropic claimed it built the program to “close the gap between educational best practices and what a teacher’s week allows.”
Meijer, however, remains skeptical, telling Straight Arrow that introducing more AI into the classroom fails to address the structural issues plaguing public education. To her, the technology represents another misallocation of public funds that the state should direct toward human staff and labor solutions.
“What this is, is essentially the Department of Education being willing to put money toward something as long as it’s not addressing the underlying labor issues,” she said.
Is AI a time saver or a Band-Aid?
Anthropic’s program allows educators to input specific curriculum goals so Claude can generate tailored lesson plans, student worksheets and quizzes. While the tech firm frames this as a way to save teachers hours of planning time, Meijer believes it strips away a fundamental element of the profession.
According to Meijer, offloading lesson design to software deprofessionalizes educators, replacing deep educational reflection with algorithmic outputs.
“I would argue that the act of writing the lesson plan is the lesson,” Meijer said. “The process of writing a lesson plan and considering the actual human beings in front of us is what matters.”
She also takes issue with the underlying dynamics of Big Tech entering public education. While Anthropic states that data submitted through Claude for Teachers is excluded from model training, Meijer and other critics remain skeptical, arguing that tech companies rely on educators’ unpaid expertise to build and refine automated tools that may eventually replace human labor.
“We’re training their models for free, and they’re telling us it’s a gift,” Meijer said. “No, no, no — our labor, our experience, our intellectual work — that is a gift. And I don’t give that gift to corporations. I give that gift to children.”
Anthropic, however, insists that Claude is not meant ot operate as an autonomous, “set-it-and-forget-it” tool. The company emphasizes that human oversight remains central to its design.
“For every sector that adopts AI, our counsel is the same: it is designed to augment human potential,” a company spokesperson told Straight Arrow. “We don’t want teachers to just set Claude to work developing plans without their involvement and agency. That’s why we’ve designed our teacher-specific workflows to keep teachers in the loop.”
While Anthropic is offering a free year of access for individual K-12 teachers, privacy advocates warn that free tiers are often a strategy to build product dependency before converting school districts into paid enterprise contracts. For a district like New York City, which employs roughly 81,000 teachers, software licensing can quickly scale into a multimillion-dollar commitment.
Cassie Creswell, an advocate with the Parent Coalition for Student Privacy, argues that public schools facing severe financial crises cannot afford to divert resources toward software automation. She pointed to Chicago Public Schools, which is currently navigating a budget deficit exceeding $700 million.
“We could instead put that money into having more adult certified licensed humans in our schools so that they would actually have the time to write lesson plans and not want to outsource lesson plans to robots,” she said.
The privacy trap
While Anthropic says a dedicated district-wide offering is in development, Claude for Teachers currently operates through individual educator sign-ups. Because teachers register independently rather than through formal district procurement, the tool bypasses school IT departments, creating immediate regulatory compliance concerns, according to Leonie Haimson, co-chair of the Parent Coalition for Student Privacy.
Primary among those concerns is the Family Educational Rights and Privacy Act, or FERPA, which prohibits schools from sharing personally identifiable student data with third parties without parental consent. Third-party vendors can access student data under FERPA’s “school official exception,” but only if the school district retains direct contractual control over how that data is managed.
Haimson emphasized that individual teacher accounts fail that legal standard.
“According to FERPA and many state privacy laws, the district needs to be in control of the data, and that generally requires a contract with legally enforceable terms,” Haimson told Straight Arrow. “Since this is only available to teachers, if a teacher signs up, that means the school and the district is not in control of the use of the data.”
Anthropic maintains that its platform is safe and legal, noting that it uses a dedicated K-12 data processing agreement and relies on teachers to omit identifiable student details. But Creswell argues that relying on manual “de-identification” provides false security.
“The truth is de-identified data is kind of not really a thing,” Creswell said. “It takes so little information to re-identify. To expect that the burden falls on teachers to be careful of what they’re putting into these is not fair to teachers, and it’s not good for students.”
Despite those concerns from privacy advocates, the tool has earned notable support from labor leadership. The American Federation of Teachers praised Anthropic’s approach, pointing to key guardrails, including plain-language privacy terms, restrictions against using student data to train future AI models and limiting access strictly to adult educators rather than students.
The AFT has also partnered with Anthropic to co-create training modules designed to help teachers integrate Claude into their workflows.
Following initial questions regarding data governance, Anthropic updated its announcement materials on Tuesday to clarify that whether and how educational data can be used in Claude for Teachers is ultimately determined by local district and state policies.
Classroom double standards
As AI continues its rapid expansion into daily life, researchers and child development experts are raising red flags over its long-term impact on human cognition. A growing body of research warns that over-reliance on generative tools can lead to “cognitive offloading” — a phenomenon where outsourcing mental effort to technology gradually weakens critical thinking, memory retention and problem-solving skills.
While Anthropic emphasizes that Claude for Teachers is strictly restricted to users 18 and older, child development advocates warn that introducing generative AI into teacher workflows inherently normalizes the technology in K-12 spaces. Jean Rogers, the director of the Screen Time Action Network at Fairplay, argues the push into schools threatens fundamental cognitive growth.
“There’s no question, generative AI products are both undermining and eroding cognitive development,” Rogers told Straight Arrow. “These products encourage what we call cognitive offloading, and they impede that effortful process. That’s reasoning, critical thinking.”
Anthropic and other tech industry leaders argue that introducing AI tools early in education prepares students for an automated workforce, where AI fluency is increasingly demanded. Rogers pushes back on that rationale, pointing out that AI interfaces are deliberately engineered to be intuitive. Because the tools are designed for immediate usability, adults can easily master them as they enter the job market without skipping essential developmental milestones during childhood.
Beyond cognitive development, introducing AI into instruction risks straining the relationship between teachers and students. Creswell argues that when educators turn to AI to generate assignments while penalizing students for doing the same, it creates a glaring double standard.
“To tell us not to use it — that it’s cheating if you’re using AI to write your assignments — and then somehow our instructors are using it, it really undermines that relational trust,” she told Straight Arrow. “It undermines the idea that everyone is engaged in something that’s effortful.”
These tools also raise equity concerns inside the classroom. In schools where AI is already deployed, Meijer notes that algorithmic bias can penalize students based on race, dialect or speech patterns. When students with speech impediments or those learning English as a second language interact with automated reading and grading tools, they are frequently misidentified or unfairly punished. Haimson argues that when tech companies frame these errors as a need for “more training data,” they are essentially treating vulnerable public school students as test subjects for flawed software.
Meijer points to a deeper systemic divide emerging between public and private education. It is hypocritical, she argues, for wealthy AI executives to enroll their own children in elite, screen-free private schools while marketing automated software to underfunded public districts as a low-cost substitute for human connection and specialized staff.
The future of education
Rogers and Fairplay believe the only responsible path forward is to halt the expansion of classroom AI entirely. The coalition is advocating for a five-year moratorium on generative AI tools in PreK-12 classrooms.
“Until any of this AI is proven safe and effective, there’s no place for it in the K to 12 classrooms,” she told Straight Arrow.
Anthropic has pushed back against calls for a formal moratorium, pointing to its pilot partnership with Detroit Public Schools as proof that AI can be deployed responsibly to support overworked staff.
“We have paused — we took the time to really engage with teachers and experts across the education ecosystem to make sure we were being as thoughtful as possible before we rolled out Claude for Teachers,” the company told Straight Arrow. “But we believe that AI has huge benefits when it can empower teachers to better support their students.”
Yet resistance is growing among parents, who are increasingly wary of algorithms managing their children’s education. Haimson noted that in her decades of education advocacy, she has rarely witnessed such cross-ideological agreement.
“Parents from the left and the right, Democrats, Republicans, parents with no political affiliation at all, hate it,” she said.
Rather than rushing to adopt unvetted software, Haimson’s coalition argues that school districts must subject any proposed AI tool to four mandatory evaluations before it enters a classroom. This includes independent research proving educational efficacy, a comprehensive privacy impact assessment, a third-party data security audit and rigorous testing for algorithmic bias.
As school boards and tech executives negotiate the future of digital literacy, public education stands at a pivotal crossroads. For educators like Meijer, the fight isn’t about resisting technological progress but about preserving the human foundation that makes learning possible in the first place.
“There’s always money for some new shiny object that is deemed to have something to do with ‘the future,’” Meijer said. “If you want to fix public education, you don’t invest in algorithms — you invest in the human beings standing in the room.”
Round out your reading
- From Obama to Trump to Biden: When networks said ‘no’ to presidential speeches.
- ‘The end of American science as we know it.’ White House funding plan sparks alarm.
- AI data centers are popping up everywhere. What exactly do they do?
- The Trump administration is accused of breaking a law built to target the KKK.
- One Republican senator is dead, another is still alive. Both are the center of conspiracy theories.
Ella Rae Greene, Editor In Chief
