24/7 Teach uses artificial intelligence across our programs, our products, and our operations. We teach AI fluency. We build AI tools for classrooms. We run AI-powered workflows behind the scenes so our team can focus on the work that requires human judgment.
That means we have a responsibility to govern AI seriously, not just for ourselves, but for the families, learners, educators, and organizations who trust us with their time, their data, and their outcomes.
This page summarizes how we do it. Partners, schools, and institutions that need the full governance documentation can request it directly.
Our governing principles
Humans stay in charge of judgment. AI accelerates the workflows around our team. It does not replace the mentors, coaches, instructors, or advisors who lead our programs. Every consequential decision affecting a learner, a family, or a partner has a named human accountable for it.
Transparency when AI is involved. When AI influences a decision that affects you, you have a right to know. Our classroom AI platform, Naomi-AI, gives teachers full visibility into every recommendation the system makes, including plain-language explanations of its reasoning.
Fairness and equity, monitored and enforced. AI systems that assess, group, or route learners are monitored for disparate impact across demographic groups. When bias is detected, the affected feature is paused for remediation before it continues operating.
Data minimization and privacy. AI systems collect only the data they need to do their job. Student data receives our highest level of protection. We do not use student data to train AI models for other customers. Full details are in our Privacy Policy.
Accountability with a name attached. Every AI system in our organization has a named person responsible for its operation. Every consequential AI output is logged in a way that allows after-the-fact review.
Safety first, especially for minors. AI systems that interact with children receive enhanced safety review before deployment. Content generated by AI for young learners is reviewed for age-appropriateness. No AI feature substitutes for the presence and oversight of a qualified adult.
Honest measurement. We do not overstate what our AI systems can do. The same standard of honest measurement we apply to our program outcomes applies to our AI claims: verified data, not assumptions.
How we classify and oversee AI systems
We maintain an AI System Registry that catalogs every AI system in use across 24/7 Teach and Naomi-AI. Each system is classified into one of four risk tiers based on who it affects, what data it accesses, and what happens if it fails.
Critical (Tier 1): Systems that directly affect minors, process student education records, influence educational placement, or touch a written guarantee. These require a full risk assessment before deployment, continuous monitoring, and a human override mechanism that teachers or staff can use at any time.
High (Tier 2): Systems that interact with adults in trust-critical contexts, process personal information, or influence financial decisions. These require human-in-the-loop review for final decisions.
Moderate (Tier 3): Systems that generate content under the 24/7 Teach brand or handle routine communications. These require human review before anything goes out the door.
Low (Tier 4): Internal productivity tools. These are documented and covered by our acceptable use standards.
The registry is reviewed quarterly and updated whenever a new system is deployed.
The line between automation and human judgment
We have a simple test we apply before automating any workflow. We call it the two-gate test.
Gate 1: Does this workflow touch a guarantee, a minor, a legal document, or protected student data? If yes, a human stays in the decision loop. AI can assist, draft, calculate, and surface context, but a person makes the final call.
Gate 2: Would a parent, learner, or school partner feel betrayed if they learned an AI agent handled this with no human involved? If yes, we keep the human in the loop and use AI to make them faster.
Both gates must be clear before a workflow is fully automated.
In practice, this means our back-office operations are heavily AI-assisted: scheduling, status communications, routine logistics, content drafting, and data processing all run with AI support. But the teaching, mentoring, coaching, admissions conversations, guarantee determinations, and partner relationships stay human-led. That is where trust is built, and trust is not something we automate.
Student data and educational privacy
24/7 Teach and Naomi-AI comply with the Family Educational Rights and Privacy Act (FERPA), the Children's Online Privacy Protection Act (COPPA), New York Education Law 2-d, and applicable state student privacy laws in every state where we operate.
When AI systems process student data in school settings, they operate under the school official exception with the educational institution maintaining control. Student data is never sold. Student data is never used for advertising. Student data from school accounts is not used to train AI models for other customers, except in de-identified and aggregated form or as explicitly permitted by a written agreement with the school or district.
For schools and districts that require formal data agreements, we execute Data Processing Addendums (DPAs), Business Associate Agreements (BAAs), and National Data Privacy Agreements (NDPAs) with the specific protections your institution requires, including 72-hour breach notification and 60-day data disposition upon request.
Our full privacy practices are documented in our Privacy Policy.
Naomi-AI: classroom-specific governance
Naomi-AI is our AI intervention and tutoring platform for grades 3 through 9 in math, ELA, and social-emotional learning. Because it works directly with children in classroom settings, it operates under our highest governance standard.
Teacher authority is final. Naomi-AI makes recommendations. Teachers make decisions. Every grouping decision, every content recommendation, and every debrief can be overridden by the teacher at any time.
Bias monitoring is ongoing. Student grouping outcomes are analyzed by demographic dimensions on a regular cycle. If any group of students is disproportionately affected, we investigate and remediate before continuing.
Content is age-appropriate and standards-aligned. AI-generated content for students is calibrated to grade-level standards and reviewed for developmental appropriateness across K-5, 6-8, and 9-12 bands.
Parents have visibility. Families can contact us or their school to learn how AI is used in their child's learning experience and to request review of AI-generated assessments or recommendations.
Our framework and alignment
Our AI Governance Policy is anchored to the National Institute of Standards and Technology AI Risk Management Framework (NIST AI RMF 1.0), the most widely referenced AI governance standard in the United States. The framework is structured around four functions: Govern (who makes decisions), Map (what systems exist and what risks they carry), Measure (how we monitor performance and fairness), and Manage (what we do when something goes wrong).
We also align with the principles of the EU AI Act, which classifies education AI as high-risk and requires enhanced oversight, transparency, and human review. While we are a US-based organization, this alignment prepares us for international partnerships and reflects our belief that the highest standard should be the default, not the exception.
Request the full documentation
The summary on this page reflects the commitments in our full AI Governance Policy. The complete document includes our AI System Registry requirements, risk classification methodology, decision authority matrix, automation boundary map, vendor assessment criteria, incident response protocols, acceptable use standards, and training requirements.
School districts, universities, organizational partners, and institutional reviewers can request the full governance documentation.
Request received.
Thanks. A member of our team will follow up with the full AI Governance Policy and supporting materials within two business days.
Related policies
Last updated: July 2026