Home / Technology / AI BSL Translations Spark Access Fears
AI BSL Translations Spark Access Fears
19 Jan
Summary
- AI BSL translation errors risk spreading misinformation.
- Deaf community input is crucial for accurate AI development.
- Concerns exist over patient diagnoses and court evidence accuracy.

Ministers have been alerted to significant risks associated with AI translations of British Sign Language (BSL), with fears that the technology could spread misinformation and negatively impact accessibility for disabled individuals. Charities and Members of Parliament have voiced concerns, particularly as many BSL users consider it their first language.
One instance highlighted involved an AI-generated BSL video from the NHS regarding measles, which had to be withdrawn due to grammatical errors that rendered the meaning entirely misleading. This underscores the potential for inaccurate information, which could have dangerous repercussions. Experts stress that involving BSL users from the outset is vital for improving service access and preventing wasted public funds.
While AI holds potential for inclusion, its development must be guided by the deaf community to ensure quality and accuracy. Organizations are advocating for a deaf-led framework for government AI-BSL spending, ensuring that those most affected are in control of the technology's development and implementation.




