Dear Mayor Lurie and Superintendent Su,

As concerned parents, caregivers, educators, and citizens, we are calling on you to enact a two-year moratorium on the use of generative AI in San Francisco schools.

This moratorium would prevent new SFUSD contracts for products or curricula that use AI, and hit pause on current use of student-facing and student data-mining AI products, including those that are embedded in curricula. A moratorium would allow you and the School Board to establish rigorous processes for responsibly assessing the risks and impacts of AI on student learning, privacy, and the environment, before introducing them in classrooms and schools.

Beyond this, it would send an important message to the world: San Francisco, the city at the heart of innovation and bleeding edge technology, populated with some of the most forward-thinking and technology-savvy people in the world, is approaching new technologies and how they affect our families with extreme rigor and thoughtfulness, and encouraging others to do the same.

We’re calling for a moratorium on AI in schools because:

AI impedes children’s ability to learn. The early childhood years are an overwhelmingly important time in human brain development. Introducing generative AI too soon has a negative effect on students’ ability to solve problems, think critically, and be creative. The RAND Corporation’s report on AI in K-12 contexts finds no teaching and learning gains from AI use. There is increasing evidence that AI use rewires your brain, causing effects similar to age-related cognitive decline within a short period of habitual use. A recent report from the Brookings Institute concludes, “after interviews, focus groups, and consultations with over 500 students, teachers, parents, education leaders, and technologists across 50 countries, a close review of over 400 studies, and a Delphi panel, that at this point in its trajectory, the risks of utilizing generative AI in children’s education overshadow its benefits.”

AI tech companies cannot deliver measurable positive outcomes. The generative AI “boom” and explosive growth of the wider EdTech industry has introduced many new, untested products to the market, often with waves of scientific-sounding PR promises about student outcomes. The truth is that most of those promises are empty, and students, teachers, and parents are already getting fed up with existing EdTech products in schools. Mayor Lurie, you have consistently championed accountability, transparency, and rigorous fiscal oversight across public and nonprofit sectors. Allowing public funds to be spent on unvetted tech contracts without proof of efficacy violates the core principle of responsible governance. A moratorium ensures that City and District funds are not underwriting commercial experiments that lack empirical backing, protecting taxpayers and students alike from wasteful and predatory technology procurement.

AI undermines the relationship between teachers, students, and families. A core component of education is the guidance, human witnessing, and relationships between students and educators. When curriculum design, assignments, classroom activities, communication with students and parents, or feedback assessments are outsourced to AI, the crucial trust between teacher and families is broken. When the “human in the loop” is no longer in charge but relegated to the position of Editor, tweaking the algorithmic output, the result is a pedagogical crisis. Communication and assessments generated by a Large Language Model can only synthesize and make guesses based on input data points. AI can thus generate a report that sounds reasonable, polished, and professional but lacks true witness, nuance, and meaning.

AI use threatens mental health and social development. American Federation of Teachers President Randi Weingarten recently called for a ban on student-facing AI tools in elementary schools nationwide in addition to prohibiting screen use among students in pre-K through 2nd grade. According to AFT, the AI ban is necessary to prevent harm to younger students and to help them develop skills to build relationships and resilience. Mayor Lurie, throughout your leadership founding Tipping Point Community, you dedicated many years to addressing early childhood support and youth mental health, clearly recognizing that human-centered support systems are the foundation of opportunity. Introducing addictive, data-mining, and isolating technologies into classrooms directly undermines youth mental health and emotional resilience, blowing up the preventative safety nets our most vulnerable citizens need to build resilience and successful lives.

AI reliance adds to the negative impact of overall technology use in young students. We already know that increased screen use, smartphone reliance, and social media use is highly correlated with depression, anxiety, sleep deprivation, cyberbullying, and isolation in children and adolescents. Now, young people have begun forming emotional attachments to chatbots, in some cases leading to tragic outcomes like self harm and suicide. Studies have demonstrated that AI chatbots can be manipulated to provide instructions for suicide, even when guardrails are supposedly in place, and multiple technology companies like Meta are involved in lawsuits with grieving parents who have lost children to chatbot-assisted suicide.

AI models are filled with racial, ethnic, and gender biases. LLMs produce results based on their limited sets of source data, and we know that the largest datasets which serve as the foundation of the most popular genAI products are filled with material that includes stereotypes, negative gender and ethnic biases, pornography, child sexual abuse material, and otherwise does not accurately or positively represent the reality of the world. This influences curricula, feedback, and outcomes for students. Because commercial AI models are built on biased datasets, they risk reinforcing systemic inequities and tracking already marginalized students into lower expectations. A moratorium safeguards true educational equity, protecting students from tools that may serve to widen the achievement gap for San Francisco’s diverse student body.

AI in schools open students to needless surveillance and privacy risks. Most technology companies mine our data, and young students are not sufficiently knowledgeable or sophisticated to be able to understand and consent to the risks. Private technology companies have no business collecting, using, training models on, or selling our students biometric data, private thoughts, search histories, images, learning patterns, voices, video, or anything else. Parents deserve full transparency and the right to opt out of invasive surveillance and data collection, without impacting students’ ability to fully participate in their education.

AI in schools runs counter to the city’s commitment to positive climate action and social equity. San Francisco promotes a public commitment and climate action strategy centered around reducing waste and encouraging buildings and vehicles to aim for 100% renewable energy. AI products are driving the expansion of fossil fuels, contributing to climate collapse. Our Climate Action strategy emphasises racial and social equity, addressing the ways in which the harms of the climate crisis disproportionately hit racial and ethnic minorities. Funneling contracts and dollars to AI products directly fuels climate collapse, the worst impacts of which are most heavily felt by already disadvantaged and historically marginalized people around the world.

We urge educators and administrators to pause, consider, and study the many negative risks posed by genAI to our children before rolling out more potentially harmful products in our schools. Technology moves quickly, and this technology is still very new. It is not too late to take a considered approach and to actively shape technology to maximize the potential benefits while protecting our children from harm to their cognitive, social, and emotional development. This includes a willingness to reconsider products already in use, like Google Workspace for Education (which is currently available to SFUSD students, and which includes Gemini AI features that urge students to let AI write for them) and Adobe Express (which is also widely used by students and includes Firefly AI features that urge students to let it create images for them).

We want to nurture, teach, and prepare our next generation to experience, create, and thrive in the world. We don’t want to teach them that they can’t write, can’t learn, and can’t create on their own. Mayor Lurie, having founded Tipping Point Community, investing in childhood development and mental health, you clearly understand that our children’s foundational years require real human care, stability, and protection, as opposed to automation and unauthorized experimentation. Mayor Lurie and Superintendent Su, we urge you both to use your leadership to help San Francisco hit pause, audit vendor relationships, and ensure that public dollars and programs are focused on genuine human-led learning.

A world devastated by climate change and inequality, where cognition is just a service to be purchased from technology companies, isn’t a world we want to give our children. Thankfully, we don’t have to.

Together, we can make sure our kids have a better future.