AI Privacy for Education

FERPA (Family Educational Rights and Privacy Act) prohibits the disclosure of student education records without consent. When an educator types a student name and grade into ChatGPT, that constitutes a disclosure to a third party — the AI provider. Most AI providers are not FERPA-compliant by default. Their terms of service typically allow data retention, model training, and storage in jurisdictions that may not meet institutional requirements. Even with data processing agreements, the safest approach is to never send real student data to AI in the first place. In the UK and EU, GDPR imposes similar restrictions on processing children's personal data. Schools and universities face increasing scrutiny over how staff use AI tools with student information.

MyYaad replaces student names, grades, ID numbers, and other personally identifiable information with realistic shadow values before any prompt reaches an AI provider: - Student names become different realistic names - Grades and scores change to different plausible values - Student IDs are replaced with shadow identifiers - Dates of birth shift to different dates - Addresses and contact details are substituted The AI receives enough context to generate useful feedback, lesson plans, or assessment materials — but with no real student data. When the response comes back, MyYaad restores the original values so the teacher sees the output personalised for the real student.

Teachers and administrators use AI for a wide range of tasks that involve student data: - Writing individualised student feedback and reports - Creating differentiated learning materials based on student performance - Analysing assessment data for patterns and interventions - Drafting communications to parents about student progress - Generating IEP (Individualised Education Programme) documentation - Summarising student records for internal reviews Each of these scenarios involves personally identifiable student information. MyYaad shadows this data automatically, allowing educators to work efficiently with AI while maintaining full FERPA and GDPR compliance.

Educators frequently work with spreadsheets of student data — grade books, attendance records, assessment results. MyYaad can parse XLSX files locally and shadow specific columns (e.g., shadow student names but keep subject and grade columns visible). PDF report cards, DOCX progress reports, and other documents can also be parsed and shadowed before being used as AI context. The original files never leave the educator's device.