Many AI detectors have become available in the marketplace. Whether free or paid, these tools collect student data and use it to grow their systems. Inputting student work into AI detection platforms without the student’s written consent can violate the Family Educational Rights and Privacy Act (FERPA).
Family Educational Rights and Privacy Act (FERPA)
Under FERPA, student work is considered an educational record and thus requires written affirmative consent from the student to share it. The only exception is when the college contracts with an outside party engaged in service to the institution, who is under the control and parameters of a contract with the institution and is bound in their own activities by FERPA regulations.
Additional situations pertaining to the criminal justice system, national emergencies, or financial aid can also constitute exceptions. AI detectors do not fit into any of those categories.
Running sentences from a student’s paper through a regular internet search engine to detect plagiarism does not constitute a FERPA violation.
Information Protected Under Local, State, or Federal Privacy Laws
Quinnipiac community members must ensure that any information input into Generative AI tools complies with applicable privacy laws such as HIPAA, FERPA, COPPA, and state regulations.
General Privacy Protections (FERPA, COPPA, etc.)
- Non-QU Approved: Do not enter personally identifiable information (PII), student records, or confidential university data into public AI tools (e.g., ChatGPT, Gemini). This violates university policy and applicable privacy laws.
- QU Approved: You may input content into approved university-supported tools only after removing all PII. For example, de-identified student work may be used in tools like Microsoft Copilot if handled securely.
HIPAA-Regulated Contexts
While Quinnipiac is not a HIPAA-covered entity, many faculty and students work in healthcare settings where HIPAA standards still apply.
- PHI should never be entered into AI tools unless explicitly approved by the host institution and configured for HIPAA-compliant use.
- If you are involved in clinical work, follow the privacy policies of the partner institution and consult IT before using AI tools in that environment.
False Positive Results
In addition to a violation of FERPA, AI detectors can produce false positives, particularly for students who are English language learners. Given the unreliability and potential legal and ethical problems these tools present, we strongly recommend that all faculty rely on other means to detect academic dishonesty and communicate with students about their suspicions.
Learn more about Academic Integrity at QU →
SafeAssign in Blackboard
Instructors may use SafeAssign, which is available through Blackboard. SafeAssign does not detect AI, but it can help to verify the originality of the assignments. This tool provides an “originality report” which highlights text that matches sources in its database. Instructors using this tool should review report details to confirm that highlighted text is quoted and properly cited.
Quinnipiac has published its AI Policy on QU.edu. Explore the policy →
Visit Another Section:
- Syllabus →
- Assignments →
- IgniteAI in Canvas →
- AI Tools at Quinnipiac →
- Academic Integrity and AI Detectors →
- Career Development →
- Additional AI Resources →