AI ethics is no longer merely a matter of broad principles for regulators, particularly with the emergence of autonomous agents capable of making decisions and taking actions with minimal human intervention. In Saudi Arabia, the Saudi Data & AI Authority (SDAIA) maintains that an intelligent system does not bear legal responsibility when harm occurs; responsibility remains with the developer and the entity operating it. Saeed bin Mohammed Al-Shahrani, SDAIA spokesman, told Asharq Al-Awsat in an exclusive interview that traditional governance frameworks are no longer sufficient to keep pace with such developments. In his view, high-risk decisions require human oversight, clear chains of responsibility and documented, operational emergency shutdown mechanisms. His comments come ahead of the fourth UNESCO Global Forum on the Ethics of Artificial Intelligence, hosted in Riyadh from Sept. 14-17, marking the first time the international forum is being held in the Arab world. More than 100 speakers will take part in dozens of sessions and workshops, including ministers, policymakers, experts, academics, representatives of international organizations and the public and private sectors, as well as technology executives. Global Forum Priorities The program begins with preparatory meetings and gatherings of international networks involved in implementing UNESCO’s Recommendation on the Ethics of Artificial Intelligence, as well as a closed strategic briefing for ministerial participants and a ministerial dialogue on international cooperation. The forum will review five years of implementing UNESCO’s recommendation, adopted in 2021, and discuss financing governance systems and building regulatory capacity. It will also see the launch of the updated Readiness Assessment Methodology, RAM 2.0, based on analysis of more than 55 readiness reports. Sessions will address practical issues including AI use in public administration, health and humanitarian work, youth mental health, the environment, energy and water, linguistic diversity — including Arabic — as well as physical AI, robotics and human-machine relations in the workplace. From Principles to Engineering Requirements Five years after UNESCO adopted its AI ethics recommendation, Al-Shahrani sees the main problem as no longer a lack of principles, but the difficulty of translating them into tools developers and regulators can use and against which compliance can be measured. He described the key gap as the “operational translation of principles”: turning concepts such as fairness, transparency and accountability into engineering requirements, standardized tests and binding regulatory controls for developers. He also pointed to disparities in governments’ ability to conduct hands-on technical assessments and monitor systems before deployment and during operation. In Saudi Arabia, SDAIA has introduced principles and frameworks governing the relationship between people, data and AI technologies, Al-Shahrani said, covering integrity and fairness, security and privacy, humanity, social and environmental benefits, reliability and safety, transparency, as well as accountability and responsibility. Flexible Regulation Al-Shahrani identified building specialized governance skills and institutional capacity as a major readiness challenge, noting that technology is evolving faster than people can adapt. He continued that SDAIA is pursuing a “flexible regulation” strategy based on continuous assessment and providing companies with self-assessment tools rather than imposing complex bureaucratic procedures. The approach comes amid growing debate within the AI industry over whether safety and oversight measures can keep pace with the rapid development of advanced models. From Adopting Standards to Shaping Them SDAIA does not necessarily expect the world to reach a single detailed AI law. Al-Shahrani acknowledged the difficulty of establishing one detailed framework given differences in legal systems and economic priorities among the United States, Europe, China and others, but considers convergence on fundamental principles and safeguards “possible and necessary.” What is needed, he argued, is not a single dominant model but common ground on issues such as safety and ethics, while narrowing the gap between technological powers and emerging markets. Al-Shahrani told Asharq Al-Awsat that Saudi Arabia has become an active participant in shaping international AI rules and initiatives by sharing its expertise, partnerships and a vision that seeks to balance innovation and governance. He highlighted cooperation with UNESCO and other international institutions, as well as the establishment in Riyadh of the International Center for AI Research and Ethics (ICAIRE), under UNESCO’s auspices. The center is intended to help move AI ethics from broad principles to practical implementation by embedding them in scientific research, education, standards, innovation and public policy from the earliest stages of technology development — reflecting Saudi Arabia’s desire to help shape standards rather than merely apply rules formulated elsewhere.
