top of page
Search

Who Gave Approval for AI with Sign Language?


AI with sign language is moving fast. Accountability is not.


Too many claims are appearing in very high numbers. Too many organisations are still avoiding a clear public position. And too often, the discussion seems more focused on protecting AI than protecting people.


That is why my question remains simple:


Who gave approval for AI with sign language?


Alt text: Bold campaign graphic on a dark background. Large white and yellow text says, “AI with sign language is moving fast. Accountability is not. Who gave approval? Where is the accountability?” A glowing blue AI face appears on the left with digital light trails. On the right, a close-up of a hand is raised in a sign, highlighting the contrast between AI and human sign language.
Alt text: Bold campaign graphic on a dark background. Large white and yellow text says, “AI with sign language is moving fast. Accountability is not. Who gave approval? Where is the accountability?” A glowing blue AI face appears on the left with digital light trails. On the right, a close-up of a hand is raised in a sign, highlighting the contrast between AI and human sign language.

Claims are growing fast, but public accountability is still far too weak. I have also noticed that too many associations, charities, and organisations in the UK still avoid giving a clear opinion on this.

That silence matters.


Silence is not leadership.


If organisations genuinely care about safeguarding, standards, ethics, and Deaf people’s rights, then they should be prepared to say clearly where they stand on AI and sign language.


For clarity, this is not a question about whether the dataset was licensed for commercial use. It is a question about whether the broadcasting firm holds any recorded information showing awareness of any unlicensed commercial use of the dataset. That is a different question, and it matters for accountability.


I have also asked through FOI because I want to understand who approved the use of AI with sign language, on what basis, and with what accountability.


That is the issue.


If organisations do not put people first, then what exactly are they safeguarding? Too often, it feels like the focus is on protecting AI, not protecting humans.


BSL is not for sale.

Deaf people are not a testing ground.

Sign language should never be treated as a product to exploit.


There is useful work already in circulation. The European Union of the Deaf published important material on artificial intelligence and sign language, including resources that help frame the discussion around ethics, governance, and protection. That is valuable, especially for EU countries.


In the UK, the report BSL is not for sale: A Deaf-led approach to AI procurement also makes an important contribution. It points towards a Deaf-led way of thinking about standards, procurement, and accountability.


But this debate is not only for the EU, and it is not only about one country. The UK and other non-EU countries need open discussion too. They need stronger leadership, clearer public positions, and a much more serious approach to safeguarding.


What is needed now is not silence, but Deaf-led standards that work across healthcare, education, justice, and other public services.


Alt text: Illustration showing Deaf communication across four settings: healthcare, education, policing, and justice. A blue circular logo in the centre reads “ISO/IEC 42001 BFI AI with Sign Languages.” Around it, Deaf people are shown signing with a doctor, teacher, police officer, and judge.
Alt text: Illustration showing Deaf communication across four settings: healthcare, education, policing, and justice. A blue circular logo in the centre reads “ISO/IEC 42001 BFI AI with Sign Languages.” Around it, Deaf people are shown signing with a doctor, teacher, police officer, and judge.

We already have ISO/IEC 42001 for AI management systems. British Sign Language also has its own recognised code, bfi. But what is still missing is a proper standards bridge between AI and sign languages in practice.


That means standards for:


  • accessibility

  • accountability

  • accuracy

  • safeguarding

  • Deaf-led oversight

  • procurement

  • human rights and public trust


This matters in real settings. It matters in healthcare, where misunderstanding can affect safety. It matters in education, where poor quality communication can affect learning and inclusion. It matters in justice, where inaccuracy can affect fairness and rights. And it matters across public services, where Deaf people should never be expected to accept lower standards just because AI is new or fashionable.


The goal should be to reduce problems and build trust through full accessibility, accountability, and accuracy.


That should be the standard.Not an afterthought.


I want to see more organisations stop avoiding this issue and give their opinions openly. Silence may be convenient, but it does nothing to reduce harm, improve standards, or build confidence.


We need:


  • Deaf-led standards

  • clear procurement rules

  • proper safeguarding

  • transparent accountability

  • respect for the many sign languages used across the world


Because there is not just one sign language. There are many sign languages, many Deaf communities, and many contexts where poor decisions can cause real harm.


So I return to the same question:


Who gave approval for AI with sign language?


And if approval is being assumed rather than clearly demonstrated, then that should concern all of us.


Further reading





No approval without Deaf-led standards.

No safeguarding without people first.




 
 
 

Comments

Rated 0 out of 5 stars.
No ratings yet
Add a rating
bottom of page