Federal Minister for Information Technology and Telecommunication Shaza Fatima Khawaja on Tuesday said artificial intelligence (AI) systems must be developed and used in full compliance with international law and the United Nations Charter, with meaningful human oversight in both civilian and security domains.
Addressing the thematic discussion on “Safe, Secure and Trustworthy AI: Interoperability and Compatibility of Approaches” during the Global Dialogue on Artificial Intelligence Governance in Geneva, the minister said AI safety was a shared global interest but stressed that the process of defining safety must be transparent, inclusive and representative of all nations.
She said safety standards and benchmarks for trustworthy AI were increasingly being shaped by a limited number of countries and technology firms through national frameworks that excluded much of the developing world.
“Safety is not a purely technical matter. Language, culture and local realities shape it,” she said, adding that standards developed without the participation of developing countries risked being imposed rather than collectively adopted.
Shaza Fatima said Pakistan supported the formulation of AI safety norms through inclusive multilateral processes under the United Nations, where all countries could contribute as equal partners.
Highlighting the importance of interoperability, she said Pakistan supported compatible AI governance approaches, as envisioned in the Global Digital Compact, noting that interoperability promoted cross-border trade, research and data flows while allowing different national regulatory systems to coexist. However, she cautioned that interoperability should serve as a bridge between governance systems rather than becoming a mechanism through which the rules of a few countries were imposed on others.
The minister emphasized that global AI standards should be developed through open and collaborative standard-setting bodies, ensuring safety, reliability, respect for human rights, and recognition of linguistic and cultural diversity throughout the AI life cycle.
She further stressed that trust in AI depended on countries having the capacity to independently test, evaluate and verify AI systems.
“The ability to assess AI systems remains concentrated in the countries where the technology is developed,” she said, warning that developing nations should not be expected to rely solely on external assurances of safety without having the means to verify them. She called for strengthening AI evaluation and governance capabilities in developing countries through shared tools, technical cooperation and regional partnerships.
Shaza Fatima said safe and trustworthy AI should not mean safety for only a few while remaining opaque to the rest of the world. She reaffirmed Pakistan’s commitment to working with international partners and the co-chairs of the dialogue to build an inclusive, transparent and durable global framework for AI governance.