AI Usage Policy

Last updated: March 2026

This AI Usage Policy defines the boundaries, expectations, and necessary human oversight required when interacting with the artificial intelligence features integrated into the MyScienceHOD platform.

1. Understanding AI Limitations

The generative AI models powering MyScienceHOD are trained on vast amounts of data but lack true comprehension, real-time factual awareness, and pedagogical intuition. Users must acknowledge the following limitations:

• Hallucinations: The AI may occasionally generate plausible-sounding but factually incorrect or nonsensical scientific information.
• Contextual Deficits: The AI does not know your specific students, their prior knowledge, or unique classroom dynamics.
• Curriculum Nuance: While aligned with general standards, the AI may miss specific localized syllabus requirements.

2. The Verification Requirement ("Human in the Loop")

MyScienceHOD is a "copilot," not an "autopilot." There is a strict requirement for human verification:

• No Blind Reliance: You must read and critically evaluate all AI-generated drafts, including intervention suggestions, questions, rubrics, lesson plans, and feedback drafts, before use.
• Fact-Checking: Scientific facts, formulas, and historical context generated by the AI must be verified against authoritative educational resources.
• Final Sign-off: The educator applying the content is solely responsible for its accuracy and appropriateness.

3. Input Data Guidelines

When prompting the AI or uploading context documents:

• Do not input Personally Identifiable Information (PII) of students during the Beta.
• Real student scripts are not processed. Software that reads handwritten answers is evaluated on fictional or redacted materials only.
• Ensure that any uploaded proprietary school materials are permitted to be processed by third-party AI models under your institution's data policies.