AI Ethics Policy

Classroom Pulse is committed to the responsible development and deployment of artificial intelligence in educational settings. We believe AI should augment educator expertise, not replace professional judgment. Core Principles: • Human-Centered Design: AI serves educators and students, never the reverse • Transparency: Clear explanations of how AI recommendations are generated • Fairness: Continuous monitoring for bias and equitable outcomes • Privacy: Student data protection is paramount • Accountability: Clear human oversight and control over AI systems Our AI features are designed to: • Support data-driven decision making • Identify patterns humans might miss • Save educator time on routine analysis • Provide research-based intervention suggestions • Generate reports and visualizations All AI-generated insights require human review and professional judgment before implementation.

We believe educators have the right to understand how AI influences recommendations. Explainable AI: • Every AI recommendation includes an explanation of contributing factors • Confidence levels are clearly displayed • Data sources are identified • Limitations are acknowledged What Our AI Does: • Analyzes behavior patterns over time • Identifies correlations between antecedents, behaviors, and consequences • Suggests evidence-based interventions • Generates progress summaries and trend reports • Flags potential concerns for educator review What Our AI Does NOT Do: • Make autonomous decisions about students • Replace professional behavioral assessments • Diagnose conditions or disabilities • Make predictions about student futures • Share individual data for AI training without consent Model Information: • We use Google's Gemini AI models for analysis • Models are accessed via secure API, not trained on your data • No student data is retained by AI providers • Regular audits ensure compliance with our policies

We actively work to ensure our AI systems treat all students equitably. Bias Monitoring: • Regular audits of AI recommendations across demographic groups • Statistical analysis for disparate impact • External review of algorithms and outcomes • Continuous improvement based on findings Safeguards: • AI never has access to protected characteristics (race, disability status, etc.) in analysis • Recommendations focus on behaviors, not individuals • Multiple data points required before generating insights • Human review required for all significant recommendations Known Limitations: • AI may reflect biases present in historical behavioral data • Limited effectiveness with small data sets • May not account for cultural context without educator input • Requires diverse implementation feedback for improvement Our Commitment: • Prompt investigation of reported concerns • Transparent reporting on fairness metrics • Collaboration with researchers and advocacy groups • Regular training updates to improve equity

Student privacy is protected at every step of AI processing. Data Protection: • All data encrypted in transit and at rest • AI processing occurs in secure, isolated environments • No student data is used to train AI models • Strict access controls limit who can view AI outputs Minimization: • Only necessary data is provided to AI systems • Student identifiers are pseudonymized before analysis • Raw data is not retained after processing • Aggregate insights preferred over individual analysis Third-Party AI: • We use Google's Gemini AI via secure API • Data is not retained by Google beyond request processing • Strict contractual protections govern data handling • Regular security assessments of AI providers Your Control: • AI features can be disabled at account level • Data deletion requests include AI-derived insights • Export includes AI analysis history • Clear audit trails of AI interactions

Humans remain in control of all decisions affecting students. Professional Judgment: • AI provides suggestions, educators make decisions • All recommendations clearly labeled as AI-generated • Easy dismissal or modification of AI suggestions • No automated actions without educator approval Required Human Review: • Behavior intervention recommendations • Pattern identification affecting student treatment • Progress reports before sharing with parents • Any insights that could influence IEP/504 decisions Escalation Paths: • Clear process to report AI concerns • Rapid response team for serious issues • Option to disable AI for individual students • Feedback mechanism to improve AI quality Training: • Educator guidance on appropriate AI use • Understanding AI limitations • Recognizing potential bias • Best practices for AI-assisted decision making

Our development practices prioritize ethics alongside functionality. Development Principles: • Privacy by design in all AI features • Thorough testing before deployment • Gradual rollout with monitoring • Easy rollback if issues emerge Testing: • Diverse test data sets • Bias evaluation before launch • Edge case identification • User feedback integration Continuous Improvement: • Regular model performance reviews • Educator feedback incorporation • Research literature monitoring • Collaboration with education experts Documentation: • Detailed technical documentation • Change logs for AI updates • Impact assessments for significant changes • Public commitment to ethical standards

Our AI is designed specifically for educational behavior support. Educational Focus: • Built for FBA and BIP data collection • Aligned with PBIS frameworks • Supports evidence-based practices • Respects educator expertise Appropriate Use: • Supporting behavior pattern identification • Generating data visualizations • Suggesting research-based interventions • Streamlining documentation Inappropriate Use: • Replacing professional behavioral assessments • Making placement decisions • Predicting student outcomes • Sharing insights without proper context Professional Standards: • AI recommendations align with BCBA ethical guidelines • Support for special education best practices • Respect for IEP team decision-making • Integration with existing professional workflows

Clear accountability structures govern our AI systems. Internal Governance: • Dedicated AI ethics review for new features • Regular audits of existing AI systems • Clear escalation paths for concerns • Executive accountability for AI ethics External Accountability: • Annual transparency reports • Third-party audits when appropriate • Collaboration with advocacy organizations • Responsive to regulatory guidance Incident Response: • Clear protocols for AI-related issues • Prompt notification of affected users • Thorough investigation and remediation • Public disclosure when appropriate Continuous Engagement: • Regular stakeholder feedback • Educator advisory input • Research community collaboration • Policy maker engagement

You have rights concerning AI use with your data. Access: • View AI-generated insights about your students • Understand how recommendations were generated • Access audit logs of AI interactions • Download AI analysis history Control: • Opt out of AI features • Request human-only analysis • Correct AI-generated errors • Delete AI-derived insights Feedback: • Report concerns about AI recommendations • Suggest improvements • Flag potential bias • Share success stories Support: • Training on AI features • Documentation and guides • Responsive support team • Regular updates on AI improvements Contact: ai-ethics@classroompulse.io

This policy will evolve as AI technology and best practices advance. Change Process: • Material changes communicated via email • 30-day notice for significant updates • Opportunity for feedback before implementation • Version history maintained Triggers for Review: • New AI feature development • Emerging best practices • Regulatory guidance changes • Stakeholder feedback • Incident learnings Our Commitment: • Continuous improvement of AI ethics practices • Responsiveness to community concerns • Leadership in educational AI ethics • Transparency in policy evolution Questions or concerns? Contact: ai-ethics@classroompulse.io Classroom Pulse 5435 N Garland Ave Suite 140-127 Garland, TX 75040 (972) 439-5845