Regulating Personal Cameras for Disabled People and People with Deafblindness: Implications for HCI and Accessible Computing
Sarah L. Woodin, Arthur Theil · 2021 · Proceedings of the 23rd International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS '21) · doi:10.1145/3441852.3476471
Summary
This experience paper examines the intersection of social policy, privacy regulation, and assistive technology design, focusing on the case of personal cameras for people with deafblindness. Drawing from the EU-funded SUITCEYES project (2018-2021), the authors — a disability studies/social policy researcher and an HCI researcher — reflect on the challenges of developing camera-based assistive technology within a rapidly evolving regulatory landscape. They interviewed 79 people with deafblindness and family members across five countries (Germany, Greece, the Netherlands, Sweden, and the UK) about their technology use, desires, and views on new technologies. The paper situates the discussion within the social model of disability, arguing that while researchers face few restrictions on what they investigate, disabled people experience many barriers in how the products of design may be accessed and used — barriers mediated significantly by social policy. The authors review the regulatory context including GDPR, the EU Artificial Intelligence Act (2021), and varying national implementations of face recognition regulation. They highlight that personal cameras with face recognition capability are particularly contested because they sit at the intersection of accessibility needs and mass surveillance concerns, with current regulations providing no specific accommodations or exemptions for people with sensory impairments who might use such technology for essential identification purposes.
Key findings
None of the 79 people interviewed had access to personal cameras for person or object recognition, though many saw face recognition as potentially very useful for understanding their social environment — identifying who is nearby, whether someone is speaking to them, or navigating public spaces independently. Participants currently relied on non-visual person identification strategies like touch (recognising handshakes, distinctive watches), smell (perfume, washing powder), vibration of footsteps, residual sight in bright light, and tactile sign language. Cost was a major barrier — face recognition technology costs thousands of pounds and is not funded by social services. The regulatory landscape is highly complex: GDPR applies across Europe but implementation varies (face recognition is not addressed in Greek law but requires consent of the identified person in German law); the EU AI Act bans real-time biometric surveillance except for law enforcement in specific circumstances; and no regulations provide accommodations for disabled users of personal cameras. Critically, disabled people have been disproportionately targeted by private companies through exploitative data practices — the browser company Brave reported that private companies were embedding surveillance on UK council websites where disabled people sought help. Some disabled people's organisations have prioritised campaigns against biometric data misuse over access to face recognition technology, viewing digital rights threats as more significant.
Relevance
This paper raises essential questions for the accessibility community about the tension between assistive technology innovation and privacy regulation. Designers of camera-based assistive tools — including visual assistance apps, wearable AI devices, and face recognition systems for blind or deafblind users — must engage with social policy frameworks that can determine whether their technologies actually reach intended users. The key recommendations are practical: privacy must be prioritised in face recognition for accessibility (a private system that respects others' privacy); safeguards against hate crime and exploitation must be built into assistive devices; technology must be cost-effective and affordable; designers must understand the processes disabled people navigate to gain access to funded equipment; and information about new devices must reach policymakers who control approved equipment lists. For accessibility practitioners, this work highlights that technical innovation alone is insufficient — without engagement with the policy landscape, promising assistive technologies may remain inaccessible due to cost, regulation, or lack of awareness among decision-makers.
Tags: deafblindness · wearable cameras · face recognition · privacy · data protection · social policy · assistive technology · ethics · AI fairness · GDPR
Standards referenced: GDPR · EU Artificial Intelligence Act