BSL must not become vague. BSL must remain clear, accurate, and protected — before, during, and after AI.
- Tim Scannell
- 11 hours ago
- 3 min read
On Sunday, 16 February, I published a blog post on “blurring the handshape.” The central point was simple:
When the handshape is blurred, the meaning is blurred.
This is not a minor technical issue. It is:
a language integrity issue,
an accessibility issue,
and a Deaf rights issue.
AI-generated signing is increasingly being presented as “accessibility.” However, sign language access is not merely visual output on a screen.
Sign language access is a conversation.
It depends on the interaction of:
clear handshape and movement,
facial expression and non-manual features,
body language and gaze,
turn-taking,
feedback,
clarification/repair,
and shared understanding.
If an organisation deploys AI signing as a one-way video, the key question is:
What happens when the Deaf person needs clarification?
For me, this is the central principle:
BSL must not become vague because technology is advancing quickly.
BSL was strong and fully formed before AI, and it must remain strong and fully respected during and after AI.
I want the right outcomes for the Deaf community.
I want protection for the quality and integrity of beautiful BSL.
I want Deaf-led standards, not vague claims of “accessibility.”
That requires:
accurate language labelling (BSL must not be mislabelled, including confusion with SSE),
Deaf-led verification before publication,
a clear human escalation route (interpreter / VRS / booking),
accountability when errors occur,
and explicit standards for quality, safety, and redress.
A practical next step: a formal liaison framework
I am calling for a formal liaison framework between Deaf organisations and sign language professional bodies to ensure coordinated, accountable action on AI-generated signing.
Proposed joint liaison model
Deaf organisations lead on Deaf rights, community impact, and access harm
Sign language professional bodies lead on language quality, teaching and assessment standards, and professional practice
Both jointly approve public standards, procurement guidance, and minimum requirements for any AI signing use
What this liaison framework should deliver
a published shared AI position
minimum standards for safe use (if used at all)
red/amber/green risk categories
mandatory Deaf-led verification
mandatory human escalation routes
reporting, accountability, and redress processes
engagement with commissioners (NHS, councils, courts, transport, and government departments)
Stakeholder communication and accountability
A further issue is stakeholder communication. In a recent discussion at the European Network of Sign Language Teachers (ENSLT), there was clear agreement that stakeholders are not communicating enough with Deaf communities about what is happening, what is being developed, and what is being deployed.
This lack of communication creates avoidable confusion, mistrust, and risk.
If decisions are being made about AI-generated signing, then the following stakeholders must communicate fully and transparently with Deaf communities:
Deaf community members and Deaf service users
Deaf organisations and Deaf-led advocacy groups
Sign language professional bodies
BSL teachers and assessors
Sign language interpreters and translators
Technology companies / AI developers/platform providers
Commissioners and procurement teams
Public service providers (including NHS bodies, local councils, courts, transport operators, and government departments)
Accessibility, equality, and inclusion leads
Regulators and standards bodies (where relevant)
Deaf communities need to know:
What systems are being developed,
where they are being deployed,
what safeguards are in place,
what risks have been identified,
and what action is being taken when problems occur.
Full communication is not optional. It is part of accountability.
I am a member of the Association of British Sign Language Teachers and Assessors (ABSLTA), and I delivered a presentation with ABSLTA members last weekend for the European Network of Sign Language Teachers (ENSLT). This reinforces for me the urgency of addressing AI-generated signing through robust standards, professional collaboration, Deaf-led governance, and transparent stakeholder communication.
Subject to further discussion and collaboration, I may also develop and present an ABSLTA framework at a later stage.
This is not anti-technology.
It is pro-language, pro-rights, and pro-accountability.
If AI is to be used in sign language access, then Deaf people must lead the standard, with strong liaison between Deaf organisations and sign language professional bodies.
Beautiful BSL deserves clarity — not blur.
BSL must remain solid before, during, and after AI.
Call to action
I am calling on Deaf organisations, sign language professional bodies, technology companies, commissioners, and public service providers to work together now to establish a formal liaison framework and publish shared minimum standards for AI-generated signing, including:
clear roles (rights + language standards)
minimum safety requirements
Deaf-led verification
accurate language labelling
human escalation routes
reporting, redress, and accountability
a stakeholder communication plan (how Deaf communities will be informed, consulted, and updated)
a public roadmap with timelines
The Deaf community needs action, not silence.
Comments