UN charter and joint lab: Islamic World League opens its resources to AI companies

The Islamic World League chose the week of high-level meetings of the UN General Assembly in New York to organize an international event titled “Artificial Intelligence for All: Faith, Human Dignity, and Artificial Intelligence”, with support and participation from the United Nations. The meeting was attended by the league’s secretary general and head of the Muslim Scholars’ Council, Sheikh Muhammad bin Abdul Karim Al-Isa, the high-level representative of the United Nations Alliance of Civilizations, Miguel Muratinos, the secretary general of the Organization of Islamic Cooperation, Ismail bin Al-Shaikh Ahmad, as well as representatives of major American technology companies, the World Bank, American universities and researchers in AI ethics.

The discussions focused on the need to broaden international participation in addressing technology ethics to include religious and civilizational perspectives and shared moral principles, while examining the content generated by intelligent systems and its potential to feed hate speech, violence and civilizational conflict driven by ideological and intellectual motives. The meeting concluded with a recommendation that the Alliance of Civilizations develop a “Covenant of Honour of the Alliance of Civilizations for AI Ethics”, and form a committee led by the alliance that includes religious leaders, experts in legislation and education, and technology-company representatives, to draft a common international framework, an initiative that Muratinos said his organization is ready to implement.

The meeting moved beyond theoretical discussion to launch institutional and field work tracks in coordination with global technology companies.Al-Isa announced that the league intends to create an artificial-intelligence laboratory in partnership with technology firms to develop software that religious, educational and humanitarian institutions can use. He also disclosed a plan to digitize a number of Arabic and Islamic sources and books and make them available to AI companies while protecting intellectual-property rights, as well as to establish a specialized unit to detect translation errors, correct them and disseminate the results widely, after observing recurring mistakes in conveying religious texts and related studies that have distorted concepts and built false perceptions in intelligent systems.

This development is directly linked to the reality of developers and data officers in the Gulf and the Arab world, where generative AI projects face a chronic dilemma of scarce, refined, high-quality Arabic data sets. Converting approved sources into digital formats ready for training opens the way for regional model-engineering teams to reduce levels of semantic hallucination, rather than relying on sparse or unreliable data randomly scraped from the internet. Establishing an independent translation-audit unit also provides a practical benchmark for engineers to evaluate imported language models and test their sensitivity to local cultural and legislative contexts.

The shift of regional initiatives from objecting to model biases to contributing precise data represents a practical change in the governance equation.This path gives technology managers and relevant institutions in the region clearer negotiating references with major model providers, and enables cultural-privacy protection by supplying reliable, vetted inputs, thereby raising the quality of local digital products and limiting the costs of correcting concepts after software agents are deployed in critical sectors.