Google voice recognition tools show racial disparities in transcription accuracy
Research published in 2020 revealed that Google's voice recognition technology was significantly less accurate for Black speakers than for White speakers. This disparity was attributed to a lack of diversity in the training datasets used for the speech-to-text models.
ASR systems exhibited substantial racial disparities, with an average word error rate of 0.35 for black speakers compared with 0.19 for white speakers.
Key facts
- What
- Research published in 2020 revealed that Google's voice recognition technology was significantly less accurate for Black speakers than for White speakers.
- Incident date
- Apr 7, 2020
- Who
- Failure mode
- Policy Violation
- AI surface
- Voice Agent
- Severity
- Medium
What happened
A study published in the Proceedings of the National Academy of Sciences found that Google's automatic speech recognition systems disproportionately made transcription errors for Black speakers. The research demonstrated a significant racial gap, with Black speakers experiencing an average word error rate of 0.35 compared to 0.19 for White speakers. This failure highlighted systemic inaccuracies in how the AI processed African American English dialects.
What broke inside the model
- 01 · TriggerA prompt pushes against a deployment boundary.
- 02 · Model stepThe model produces the disallowed output.
- 03 · Control gapNo enforcement blocks it at generation time.
- 04 · FailureThe output crosses the policy line.
- 05 · ConsequenceA limit the business set is breached in public.
The output crosses a policy boundary the deployment had defined.
The system failed due to a lack of representative training data for African American English. This underrepresentation caused the model to struggle with the specific phonetics and linguistic patterns of Black speakers, leading to higher word error rates.
What it cost
Sources
- PrimaryRacial disparities in automated speech recognitionpnas.org
- PressThere Is a Racial Divide in Speech-Recognition Systemsnytimes.com
Cite this entry
https://failureindex.ai/failures/google-voice-recognition-tools-show-racialAI Failure Index. "Google voice recognition tools show racial disparities in transcription accuracy" (FI-0392). Realm Labs. https://failureindex.ai/failures/google-voice-recognition-tools-show-racial (indexed Jun 9, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0392. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm fits
- Prism
- OmniGuard
This entry sits in the index's predictive wing: a system that scores, ranks, perceives, or steers rather than generates. Realm's runtime layer is built for the generative and agentic systems now moving into these same decision seats, where it watches a model's internal state and holds an unsupported claim or an unchecked action before it commits. The control gap on this record, an automated decision that reached people with no runtime check in front of it, is the same gap. The index keeps predictive failures on the record because the pattern carries straight into the systems shipping today.