Policy Briefing
Volume 6 Issue 6
Dr. Rukiya Deetjen-Ruiz*, Oscar Wendel, Dr. Lisa Cameron3 and Jacqueline Winstanley
May 28, 2025
DOI : 10.56831/PSEN-06-204
Abstract
The Global Stratalogues Roundtables, held at the UK Houses of Parliament, convened international leaders across government, technology, academia, and civil society to explore critical intersections between artificial intelligence (AI), accessibility, and inclusive infrastructure. Discussions centered on the ethical deployment of AI, the role of inclusive design in digital systems, and strategies for embedding accessibility from the outset of technological innovation. The roundtables produced clear consensus: accessibility must no longer be treated as an ancillary goal, but as a foundational principle guiding the governance and application of AI systems. This briefing outlines key insights and policy recommendations derived from the proceedings, with specific emphasis on scalable solutions for governments, institutions, and technology providers.
Context and Problem Definition
The accelerated deployment of AI across public and private domains has outpaced existing regulatory and design frameworks, resulting in systemic exclusions for persons with disabilities and digitally underserved populations. Despite the increasing recognition of digital rights and the importance of inclusive systems, most AI infrastructure—spanning education, finance, health, and urban mobility—remains inaccessible to individuals whose identities fall outside of majority data representation.
Emerging challenges include algorithmic bias, opaque decision-making, exclusionary data practices, and the absence of participatory design mechanisms. The problem is not one of technological incapacity, but of governance structures that fail to mandate inclusion, ethical audit, and accountability in AI development. Inaccessible AI not only deepens the digital divide, but exacerbates pre-existing inequalities across education, income, and mobility.
Key Insights from the Global Stratalogues
Roundtable participants reaffirmed that accessibility must be embedded from the earliest stages of design and development. Retrofitting systems for compliance often results in partial and ineffective access. Accessibility should be operationalized as a continuous, iterative process aligned with universal design principles.
Inclusive innovation requires the active participation of persons with disabilities in all stages of AI design. Lived experience offers non-substitutable insight into real-world usability and barriers, challenging the assumptions embedded in conventional testing frameworks.
AI systems trained on non-representative datasets risk replicating or amplifying societal biases. Data practices must prioritize diversity and contextual sensitivity. This includes strategies for capturing underrepresented identities, especially those at the intersections of disability, race, language, and socioeconomic status.
Technologies such as digital twins, AI wearables, and assistive robotics hold significant promise when applied to accessibility contexts. Adaptive interfaces, contextual recognition, and predictive support can enhance autonomy and quality of life for users with varying access needs.
Traditional, prescriptive regulatory models are insufficient to address the velocity of AI innovation. Outcome-oriented, principles-based frameworks that incorporate ethical oversight and sandbox experimentation were recommended as preferred models.
Policy Recommendations
Governments should introduce regulatory instruments requiring AI developers to conduct Accessibility Impact Assessments prior to deployment. Modeled after environmental and data protection assessments, AIAIs would evaluate systems for usability across diverse populations and propose mitigation strategies.
Public-private partnerships and cross-institutional research initiatives should be supported through targeted grants and innovation funds. Inclusive AI labs should be established to pilot co-designed solutions and conduct longitudinal evaluations of AI systems on inclusion outcomes.
National digital strategies must codify fundamental rights such as algorithmic transparency, the right to contest AI-driven decisions, informed consent, and equitable access to adaptive technologies. These principles should guide both procurement and innovation mandates across government agencies.
Public agencies should be required to integrate accessibility and equity metrics into procurement frameworks for digital systems. Vendors must demonstrate compliance with international accessibility standards and provide documentation of inclusive design practices.
Digital literacy programs must prioritize accessible, community-driven models that empower persons with disabilities to understand and shape the technologies that govern their lives. Curriculum should emphasize algorithmic awareness, rights-based frameworks, and participatory innovation.
Conclusion
The Global Stratalogues roundtables advanced a vision of inclusive technological futures grounded in accessibility, dignity, and participatory design. As AI systems increasingly mediate access to education, employment, finance, and civic life, it is essential that accessibility is treated not as a feature but as a design imperative and policy obligation. The policy recommendations presented here offer a roadmap for transforming existing digital governance frameworks to align with principles of equity, inclusion, and human-centered innovation.
Governments and institutions must now take decisive steps to operationalize these insights through regulation, funding, and cross-sector coordination. The future of inclusive AI will not be determined by technological capacity alone, but by political will and ethical intent. The next stage of global dialogue—commencing in Venice—offers an opportunity to shift from theory to action.