On 30 November and 1 December, Ljubljana hosted the international event of the European Network for National Institutions (ENNHRI) on the impact of artificial intelligence on human rights, "Co-Lab: AI - ENNHRI Capacity Building Event on Artificial Intelligence". The event brought together a wide range of representatives from European National Human Rights Institutions from more than 30 countries, as well as other international experts, to share insights and experiences, and to discuss how we can work (together) towards a future in which AI not only accelerates technological progress but also preserves and enhances fundamental rights.
At the event, Jerneja Turin, Advisor at the Ombudsman's Centre for Human Rights and Chair of the ENNHRI Working Group on the Rights of Persons with Disabilities presented some of the opportunities and challenges that artificial intelligence, automated decision making and machine learning technologies pose for the rights of persons with disabilities.
Opportunities for greater inclusion and independent living
AI holds significant potential to positively impact the lives of persons with disabilities by addressing various challenges and providing innovative solutions. From speech-to-text applications to vision-enhancing tools, AI can break down barriers and provide new avenues for communication and interaction.
AI can enhance the accessibility of digital platforms and services. Voice recognition, gesture control, and other AI-driven features can make technology more user-friendly for individuals with diverse abilities, promoting an inclusive digital environment. Navigation systems powered by AI assist individuals with visual or mobility impairments in navigating public spaces.
AI can provide tailored educational materials and support for students with learning disabilities. Adaptive learning platforms powered by AI can adjust the pace and content to meet individual needs and foster a more inclusive learning environment. Furthermore, AI technologies can create more inclusive workplaces by providing tools that accommodate diverse abilities.
If used to diagnose illnesses and recommend treatments, AI-enabled systems can contribute to better outcomes for persons with disabilities. AI also enhances the functionality of prosthetic limbs and exoskeletons, allowing for more natural and intuitive movements. Robots and other tools driven by artificial intelligence are being introduced into homes to offer care and various forms of assistance to persons with disabilities who want to live independently.
These applications demonstrate the diverse ways that AI can enhance accessibility, promote independent living and foster inclusion for persons with disabilities.
Challenges and risks
While AI provides opportunities for advancing disability rights, it is also associated with significant risks for persons with disabilities. In his report, Special Rapporteur on the Rights of Persons with Disabilities, Gerard Quinn, highlighted some.
For example, in certain circumstances, AI and emerging technologies have become gatekeepers for social protection benefits. Eligibility for benefits, such as unemployment assistance, can be restricted by AI using biased data sets and discriminatory algorithms.
In job recruitment, where AI’s use is increasing, these tools were supposed to make identifying suitable candidates more efficient and less vulnerable to human bias. However, evidence indicates that the data sets on which the machine is learning are already “infected” with bias. Historically underrepresented communities, such as persons with disabilities, are likely missing in the history of the company’s successfully recruited candidates. Therefore, they are missing in the data sets. This means that the machine will not see them as suitable candidates for the positions. In addition, chatbots used for interviews may be inaccessible to someone using a screen reader.
Faulty AI risk assessments based on certain disability types can also lead to improper increases in health insurance premiums. The lack of transparency as to the underlying logic used by machine learning and automated decision-making poses significant issues for the general population. Moreover, it is likely to impact persons with disabilities who are already marginalised in health and life insurance markets. Special Rapporteur Gerard Quinn cautioned about the gravity of these risks and the possibility that they outweigh the benefits if urgent action is not taken.
Tackling AI’s (potential) negative impacts on persons with disabilities
So, what can we do to mitigate the (potential) harms and address the challenges that AI’s increasing use poses for the rights of persons with disabilities?
The Special Rapporteur provided several recommendations to states. These include embedding human rights principles in AI regulations; explicitly prohibiting discriminatory and harmful AI use against persons with disabilities; ensuring that digital inclusion strategies consider the need for human rights-compliant AI tools; considering imposing a moratorium on high-risk AI systems, especially those with potential for discrimination; ensuring comprehensive human rights due diligence legislation that integrates disability considerations; and insisting on the obligation of reasonable accommodation in the operation of AI systems.
Among others, businesses and the private sector should operate with transparency and provide information about how AI systems work. They should implement disability-inclusive human rights impact assessments of AI to identify and rectify its negative impacts on the rights of persons with disabilities. In addition, they should actively consult persons with disabilities and their representative organisations when building technical solutions, doing so from the earliest moments in the business cycle.
NHRIs and strengthening rights in the digital age
NHRIs can play a vital role in ensuring that digital technologies strengthen human rights rather than undermine them. Using their mandates, NHRIs can address the human rights impacts of AI through diverse approaches. This includes dealing with complaints; educating and raising awareness on its human rights implications; carrying out investigations; and contributing to ongoing legislative and policy processes at both national and regional levels.
In 2023, ENNHRI and European NHRIs have been advocating for a rights-based approach to AI regulation as part of the two major, ongoing legislative processes in Europe. These are the European Union’s Artificial Intelligence Act and the Council of Europe’s Framework Convention on AI, human rights, democracy and the rule of law.
