Access Without Disclosure
Current telecom customer services often make access conditional on a user’s ability to navigate high-pressure, real-time verbal interactions. This project proposes a structural accessibility framework that removes the need for personal disclosure or diagnosis, in line with calls to move beyond purely individualised accommodation models.
By reframing customer service as an interdependent decision system, the project shifts the burden of accessibility from the individual to organisational design. This approach echoes contemporary work on universal design, systemic inclusion, and neurodiversity-informed accessibility, while remaining applicable to cognitively diverse users more broadly.
Standard telecom service models prioritise speed, persistence, and verbal fluency. These priorities inadvertently create barriers for cognitively diverse users, including neurodivergent people, who are already disproportionately excluded by neuronormative digital systems.
Existing customer service systems commonly:
Force repetition: Users must explain their issues multiple times across different channels, increasing cognitive load and “access labour”.
Mandate disclosure: Users often need to disclose personal information or disabilities in order to be taken seriously or escalated to appropriate support, reproducing well-documented patterns of stigma and risk associated with disclosure.
Create cognitive friction: Design choices related to memory, time, and authority function as structural barriers rather than technical ones, mirroring how AI-mediated services embed assumptions about attention, communication, and decision-making.
These features produce inequitable outcomes while increasing service costs and escalation rates.
This project delivers a design model and decision logic for AI-supported customer services, aligned with emerging scholarship on cognitively accessible AI and universal design in socio-technical systems.
The framework focuses on three dimensions:
Memory design
Persistent, portable context across channels and modalities to eliminate forced repetition and reduce access labour.
Time and attention design
Asynchronous service options, pause-and-return functionality, and reduced urgency cues to prevent exclusion based on processing speed, fatigue, or sensory overload.
Decision design
Transparent service pathways, lower escalation thresholds, and human-in-the-loop safeguards so that access to appropriate support is automatic rather than negotiated through disclosure or performance of distress.
The framework is explicitly designed to inform AI-mediated customer service systems by specifying when and how service logic should adapt to cognitive variability, without relying on personal disclosure or sensitive data.
This project operationalises the European Accessibility Act (EAA) by translating accessibility principles into concrete service decision logic, rather than treating accessibility as a reactive add-on triggered by individual requests.
Its innovation lies in a structural shift:
From individual disclosure to organisational design
Accessibility is embedded within service logic and AI governance, positioning cognitive accessibility as a core design condition rather than an exception.
Scalable and compliant by design
The framework can be implemented within existing AI and digital customer service infrastructures without collecting sensitive personal data, complementing GDPR and emerging AI Act safeguards while addressing their current neglect of cognitive accessibility.
For users
Reduced cognitive load and access labour, elimination of stigma and risk associated with disclosure, and more equitable customer service outcomes for cognitively diverse users.
For telecom providers
Lower escalation and complaint-handling costs, reduced churn, and more robust, fair service systems that demonstrably meet EAA and AI-ethics expectations around inclusion and public value.
Status
Conceptual framework and design model, grounded in current research on neurodiversity, inclusive AI governance, and universal design.
Team
Individual submission; research-led and design-focused, with potential for participatory co-design with neurodivergent stakeholders in later phases.