Is No One Truly Left Beh(A)Ind?: A Comparative Approach to Regulating Artificial Intelligence at the Benefit of Persons with Disabilities

In an era where data is the Twenty-first century equivalent of gold, worth and utility wise, regulating its usage, particularly when applied to today’s buzz-word—Artificial Intelligence (AI)—and tomorrow’s promise, becomes of paramount importance in guaranteeing an adequate level of protection to the digital population. In such regard, the global landscape has witnessed a dichotomous reaction to the emergence of AI solutions. On the one hand, enthusiasts have embraced AI in light of its innovative stance in shaping its surroundings. On the other, skeptics have capitulated to clichéd and, at times, slightly fictionalized prophecies emphasizing the drawbacks of AI without due account of its benefits. This difference in vision translates to disparate approaches to AI regulation, as exemplified by the European and American stances. Whereas Europe has preferred adopting a conservative and robust regulatory stance toward AI, the New Continent, in line with the vision of its Founding Fathers, has favored regulatory laissez-faireism in an attempt to reap the benefits of rapid market proliferation and implementation of AI solutions. Undoubtedly, both approaches possess benefits and pitfalls, especially for one, at times overlooked, category of users: persons with disabilities. In such regard, AI’s potential in bridging societal gaps, especially by pushing the boundaries of assistive technologies in favor of persons with disabilities, is undeniable. However, in order for AI to be a propeller for societal good, it has to be adequately crafted and duly regulated. This aforementioned potential vis-à-vis AI solutions is precisely the point of inception of this research. The comparison between the European and the American approach serves as a conceptual depiction of two polar opposite AI frameworks with intertwined objectives, namely improving the quality of life by ensuring competitiveness, attracting investments and attempting to favor innovation. However, what emerges from the research is the fact that neither system at face value appears to be the gold standard in achieving the above stated objectives, particularly with regard to persons with disabilities. The European approach elucidates various best practices in terms of guaranteed protections which are pillars to the building of a successful and trustworthy AI ecosystem at the advantage of persons with disabilities. However, such an approach lags in terms of attractiveness in favor of investments, which places Europe in somewhat of an uneasy position in attracting the necessary investments to fulfil its ambitions. On the other hand, the American stance champions innovation and investments, which in turn leads to an ever-flowing market of possibilities but can also lead to solutions that are not always conducive to the well-being and inclusion of persons with disabilities. Hence, in an attempt to enhance the European AI landscape, especially for the benefit of persons with disabilities, this research proposes a hybrid regulatory model—referred to as a ‘regulatory dalmatian’ approach—aimed at strategically merging the «best of both worlds». Such an approach is rooted in the prospect elucidated by Article 57 of the AI Act, which offers the opportunity of utilizing regulatory sandboxes at the advantage of innovation and experimentation. This approach to AI would lead Europe to retain its general regulatory landscape coated in protections and safeguards, especially for persons with disabilities, whilst allowing Europe to flirt with the American approach by integrating specks—best practices—of it into its own framework. Ultimately, reaping the benefits of the two aforementioned stances devoid of the regulatory burden that characterizes the European approach and of the dangerous-leaning laissez-faireism that dictates the American stance to AI regulation.