Skip to content
IEP Champion
Legal

AI Disclaimer

Publisher: Angel Reyes Consulting LLC. Effective: 2026-04-26 · Last updated: 2026-04-26

What the AI assistant is

The IEP Champion AI assistant is an informational tool that runs entirely on your device. It can help you draft language, surface relevant context from records you created, suggest SMART-goal phrasing, and explain common IEP-related terminology in plain language.

What the AI assistant is not

  • It is not a licensed attorney, and its output is not legal advice.
  • It is not a licensed medical or mental-health professional, and its output is not medical advice.
  • It is not a credentialed special-education advocate, evaluator, or therapist.
  • It is not affiliated with your school district, the U.S. Department of Education, or any state agency.

Limitations you should expect

  • Hallucinations. Like all language models, the assistant can confidently produce plausible-sounding statements that are inaccurate. Verify any factual claim — citations, statutes, deadlines — against authoritative sources.
  • Out-of-date training. The model has a knowledge cutoff and may not reflect recent regulatory or case-law developments.
  • No external verification. The assistant cannot look anything up online — that's by design (privacy) but means it cannot confirm a current statute, deadline, or district policy.
  • Jurisdiction blindness. IEP and special-education law varies by state, district, and school. The assistant's general guidance may not apply to your specific jurisdiction.

How to use the assistant safely

  1. Treat its output as a starting point, not a final answer.
  2. Verify factual claims against your district's published policies, your state's special-education statutes, and IDEA / 504 federal guidance.
  3. Consult a special-education attorney, parent advocate, or licensed evaluator for decisions with legal, medical, or educational consequences.
  4. Do not rely on the assistant for time-sensitive deadlines (response windows, request-for-evaluation timelines, due-process clocks). Confirm deadlines with your district or counsel.

If you spot an error

Email angel@subthesis.com with the specifics. We can't promise individual replies, but we read every report and roll corrections into future model and prompt updates.

Active Consent + Legal-Intent Filter

In response to 2026 case law on consumer-AI liability — most notably Moffatt v. Air Canada, Mata v. Avianca, the Florida Fourth District Court of Appeal's "AI slop" opinion (March 2026), and the UK Competition and Markets Authority's "Compliance by Design" guidance — passive disclaimers are no longer adequate to shield a developer from negligent-misrepresentation claims when a consumer reasonably relies on AI-generated output. The App therefore implements two structural mitigations:

  • Legal-intent filter. Free-form chat prompts containing unauthorized-practice-of-law trigger words (lawsuit, sue, due process, attorney, statute, subpoena, hearing, complaint, demand letter, etc.) are intercepted before reaching the model. The user receives a hard-coded refusal pointing them to a licensed special-education attorney.
  • Active Consent gate. Three high-risk save / export points (advocacy-letter export, Prior Written Notice response save, Document Extractor "Save selected" → IEP populate) are gated by a modal that requires the user to read the limitation copy, tick a mandatory checkbox acknowledging responsibility, and click Continue before the action fires. Consent is in-flow at the moment of action, not housed in a static document elsewhere.

Cross-references

See also our Privacy Policy (the assistant runs on-device; no prompts are sent to a server) and our Terms of Service (limitation of liability and dispute-resolution provisions).