EUD Artificial Intelligence and Sign Language

The Informed Consent & Assent Framework was embedded into three groundbreaking resources designed to safeguard the linguistic, cultural, and digital rights of deaf signers in the era of artificial intelligence (AI). Together, these materials provide a practical and ethical roadmap for engagement with AI technologies involving sign languages.

1. Template Contract for Signers and Contractors
EUD has created a model agreement to be used by deaf signers when approached by developers, researchers, or companies seeking to record and use sign language data. This contract ensures voluntary and informed consent, limits misuse of personal data and likeness, and guarantees fair conditions and remuneration. It is aligned with GDPR and EUD’s broader ethical principles.

2. Ethical Framework on AI and Sign Language
Developed in collaboration with Prof. Dr Filipe Venade, this 15-principle framework sets out the ethical foundations for AI technologies involving sign language. It calls for co-creation with deaf communities, respect for cultural identity, robust data protection, and inclusive design grounded in Deaf Digital Law.

3. New Book by Prof. Dr Filipe Venade: Sign Language in the Era of Artificial Intelligence
This publication brings together legal, academic, and community insights into how AI is reshaping sign language accessibility. It introduces the concepts of Deaf Tech and Deaf Digital Law, analyses international legal frameworks, and concludes with a declaration of ethical principles. This book is essential reading for developers, policymakers, and advocates working on inclusive AI.

These resources are free to access and designed to empower deaf communities, while guiding all stakeholders toward a future where AI serves inclusion—not substitution.

https://eud.eu/new-eud-publications-on-artificial-intelligence-and-sign-language/

The Informed Consent & Assent Framework and its accompanying resources provide a suite of benefits that are critically important for ethically engaging with deaf signers in the development of AI technologies. By embedding legal, ethical, and cultural protections into practical tools, this framework ensures that the rights and agency of deaf communities are upheld throughout the data collection and AI development process.

One key benefit lies in empowerment and protection: the Template Contract for Signers and Contractors offers a concrete mechanism for signers to assert control over how their likeness and data are used. It empowers individuals by clearly stipulating consent terms, remuneration, and usage limitations, all within the boundaries of GDPR and ethical norms. Meanwhile, the Ethical Framework on AI and Sign Language serves as a foundational guide for developers and researchers, outlining principles such as co-creation, cultural respect, and robust data protection, ensuring that AI systems are built inclusively and responsibly. Finally, the book by Prof. Dr Filipe Venade contextualizes these efforts within broader legal and technological discourses, introducing the concepts of Deaf Tech and Deaf Digital Law, and emphasizing a rights-based approach to AI. Together, these tools help prevent exploitation, foster trust, and build an inclusive digital future where AI is an enabler of accessibility, not a threat to it.

The version on OECD is out of date. EUD used a later version.
The Informed Consent framework is necessary. It is not sufficient.

When using the Informed Consent & Assent Framework and its associated tools in similar contexts—particularly where marginalized communities interact with emerging technologies—it's crucial to prioritize co-creation and cultural competence from the start. One key learning is the importance of developing legal and ethical instruments with the community, not just for them. This ensures that consent is truly informed and reflects the lived realities of participants. Additionally, integrating frameworks like GDPR and culturally specific rights (e.g., Deaf Digital Law) strengthens both the legal defensibility and ethical integrity of the engagement. For those applying this tool in other domains (e.g., indigenous languages or disability communities), the advice is to tailor template contracts and ethical principles to respect local norms, involve community representatives at every stage, and ensure accessibility—not just in language, but in format and process. Ultimately, the tool’s success hinges on viewing consent not as a checkbox, but as a dynamic process grounded in trust, clarity, and fairness.

 

It is important to be aware of the body of work by Deaf academics in this area. Of particular relevance is Buying BSL AI: Embedding Deaf Governance by By Dr Louise Hickman (Minderoo Centre for Technology and Democracy), Dr Alexa Hagerty (Minderoo Centre for Technology and Democracy), and Rose Powell (Newcastle University)