New Mexico schools push back on AI voice data policies
The gist
New Mexico school districts are battling AI reading tool Amira over student voice data privacy, forcing sweeping policy changes and legislative scrutiny.
What to know
- By mid-2026, districts like Albuquerque and Las Cruces demanded Amira delete stored student voice recordings and began offering paper-based test alternatives.
- New state opt-out policies now let families refuse voice data collection, and over 30 lawmakers are pushing for mandatory parental consent and ongoing AI oversight.
- Districts such as Albuquerque Public Schools are enforcing strict deletion protocols—wiping new voice data within 48 hours—to rebuild public trust amid heated privacy debates.
Community Backlash Spurs Change
Parents and educators across New Mexico are challenging AI reading tools, citing data privacy fears, equity concerns, and doubts about assessment accuracy—forcing districts to rethink their approach.
Parents and educators across New Mexico are challenging AI reading tools, citing data privacy fears, equity concerns, and doubts about assessment accuracy—forcing districts to rethink their approach.
Districts Hit Pause Amid Uncertainty
School leaders in Santa Fe, Los Alamos, and beyond halted AI voice assessments until the state clarifies data policies, spotlighting a growing demand for transparency and local control over student information.
By mid-2026, mounting parental concerns about the privacy of student voice data collected through the Amira AI reading assessment prompted districts like Las Cruces Public Schools to disable the retention of these recordings and request deletion of past data. LCPS also began offering paper-based assessments upon parent request, emphasizing compliance with federal and state privacy laws while actively monitoring community feedback to balance technological innovation with privacy protections.
In Albuquerque Public Schools, parental opposition intensified due to a lack of notification about Amira’s use and skepticism over the tool’s effectiveness and data handling. Parents criticized the district and state education department for implementing a convoluted opt-out process, which they argued could exacerbate equity issues by creating barriers to accountability, echoing concerns from longstanding educational equity rulings like Yazzie-Martinez.
Skepticism about Amira’s assessment accuracy grew as teachers and administrators in Maxwell Municipal Schools observed that the AI sometimes scored students based on speech patterns rather than true reading ability, disproportionately affecting those with speech delays or disabilities. Superintendent Amy Roble highlighted the district’s lack of control over Amira’s servers, raising fears of potential data breaches and long-term data misuse, which ultimately led Maxwell to discontinue the AI reading assessment despite statewide mandates and revert to the I-Ready platform to avoid further privacy risks and additional costs.
The privacy and accuracy concerns voiced by parents and districts in Maxwell, Las Cruces, and Albuquerque reflect a broader regional unease across New Mexico, with similar scrutiny emerging in Los Alamos and Santa Fe. This collective pushback underscores the growing demand for transparency, data control, and equitable assessment methods as communities grapple with the implications of integrating AI technologies like Amira into K-12 education.
Lawmakers Demand Parental Consent
A coalition of over 30 legislators is pushing for mandatory parental sign-off and real-time oversight of AI tools in classrooms, warning that sensitive student data is at risk without stronger protections.
In response to mounting privacy concerns over student voice data collected by the AI-powered Amira reading assessment, several New Mexico school districts, including Santa Fe and Los Alamos, proactively paused its use in late July 2026 while awaiting clearer guidance from the New Mexico Public Education Department (PED). District leaders, such as Los Alamos Superintendent Jennifer Guy, emphasized the necessity of transparent and verifiable policies on voice data handling before resuming use, underscoring a broader demand for accountability and control over sensitive student information.
By mid-August 2026, New Mexico had formalized policies allowing districts to opt out of using Amira’s voice-recording features, with alternatives like i-Ready and DIBELS offered to ensure compliance with state reading assessments without compromising privacy. Districts such as Las Cruces Public Schools responded to parental concerns by disabling the retention of voice recordings and providing paper-based assessment options, while Maxwell Municipal Schools discontinued Amira for K-2 reading assessments altogether, citing lack of control over data servers and potential breaches, opting instead to rely on existing platforms like I-Ready for reading and Amira ISIP for math, which do not require voice data.
Albuquerque Public Schools exemplified a comprehensive district-level approach by deleting all prior Amira student voice recordings and instituting a policy to delete new recordings within 48 hours, while also offering families the choice to opt out of the computer-based assessment in favor of a paper-and-pencil version. The district further clarified distinctions between Amira’s assessment tool, which temporarily records voices, and its tutoring platform, which does not, reflecting efforts to enhance transparency and build trust amid ongoing debates about AI’s role and oversight in K-12 education.
The New Mexico Public Education Department has empowered districts and charter schools with significant discretion over the use of AI tools like Amira, including the ability to request deletion of recorded audio, control data retention policies, and select from assessment formats that do not retain student voice recordings. This decentralized approach, coupled with the availability of alternative assessments and opt-out provisions, illustrates a state-level policy evolution aimed at balancing technological innovation with robust privacy protections and parental choice.
State Officials Signal Collaboration
New Mexico’s education department is moving toward joint action with lawmakers and parents, marking a pivotal shift toward balancing AI innovation with rigorous privacy safeguards.
In August 2026, over 30 New Mexico lawmakers, led by Rep. Linda Serrato, called on the Public Education Department to mandate parental consent and enhance transparency regarding the use of Amira AI in K-2 classrooms. Their letter highlighted serious privacy concerns about the sensitive student data collected—ranging from names and birthdays to demographic details—and stressed the necessity of informing parents about the technology’s deployment. While acknowledging AI’s potential to improve educational outcomes, the lawmakers insisted that such benefits must be balanced with rigorous oversight and clear communication to protect students and families.
The lawmakers’ demands extended beyond initial deployment to emphasize continuous monitoring for 'model drift,' recognizing that AI systems like Amira can evolve unpredictably once exposed to new data. This concern underscores the need for ongoing quality guardrails to ensure the AI’s behavior remains safe and effective over time. Parent groups echoed these worries, with voices like Lisa Marchiondo expressing frustration over the lack of awareness and consent around voice recordings and questioning the accuracy of the AI tool, which further fueled calls for stronger protections and transparency.
Responding to these mounting pressures, the New Mexico Public Education Department expressed openness to collaboration, with spokeswoman Janelle Taylor García affirming the agency’s commitment to partnering with legislators to develop clear safeguards, accountability measures, and transparent communication strategies. This willingness signals a pivotal moment in the evolving regulatory landscape, where educational authorities and policymakers are beginning to align on the necessity of balancing AI innovation with robust privacy and oversight frameworks.