Zoom's Accessibility Gaps: Captioning Accuracy and ADA Compliance Shortfalls
Deaf and hard-of-hearing users face persistent barriers as automated captioning falls short of accessibility standards
Zoom's automated captioning features, while a step toward accessibility, continue to fall short of the accuracy levels needed to provide equitable access for deaf and hard-of-hearing users. Independent testing by accessibility researchers has found that Zoom's live captioning achieves word accuracy rates of approximately 70 to 85 percent, depending on factors including speaker accent, audio quality, technical vocabulary, and background noise. While this represents a significant improvement over early implementations, accessibility experts note that accuracy below 95 percent renders captions unreliable for users who depend on them as their primary means of following a conversation.
The accuracy gap is particularly pronounced in the contexts where Zoom is most consequential. Medical consultations conducted via Zoom may involve technical terminology that automated captioning systems consistently mishandle.
Key Takeaways
- Zoom automated captioning achieves 70-85% accuracy, well below the 95% threshold considered reliable for deaf users
- Human CART captioning costs $100-200 per hour, creating a two-tier accessibility system based on resources
- Accessibility features are not uniformly available across all Zoom platforms and subscription tiers
