Artificial intelligence: Peter G. Kirchschläger advises G20
Ethics Professor Peter G. Kirchschläger and Shamira Ahmed, founder and director of the Data Economy Policy Hub in Cape Town, South Africa, have submitted proposals for the governance of artificial intelligence (AI) to the preparatory committee for this year's G20 summit in Brazil. They advocate for the establishment of a dedicated agency based at the United Nations (UN).
The G20 summit will take place in Brazil in November 2024. The substantive preparations for the discussions between the leaders of 19 participating nations, along with representatives from the European and African Unions, are well underway. Shamira Ahmed and Peter G. Kirchschläger have contributed to these preparations by submitting a joint policy paper titled “Governing Global Existential AI Risks: Lessons from the International Atomic Energy Agency”. Their proposals are part of the T20, the official think tank network supporting the summit.
AI and ethics: harnessing opportunities, mitigating risks
AI presents ethical opportunities and risks, as Ahmed and Kirchschläger explain in their document. They assert that the term “AI” is more appropriate than “data-based systems” (DS), as it better captures the capabilities of these technologies – specifically, their ability to generate, collect and evaluate vast amounts of data and act upon it. They argue that, to foster a sustainable and humane future, the regulation of AI is essential to harness its positive ethical potential while mitigating or eliminating its negative consequences.
A key measure proposed by Shamira Ahmed and Peter G. Kirchschläger is so-called “human rights-based DS”. A human rights-based approach to DS ensures that human rights are respected, protected, implemented and fulfilled at every stage of the DS life cycle and value chain, covering everything from raw material extraction and design to development, manufacturing, distribution, and the use or restriction of DS where human rights are at stake.
To support and implement human rights-based AI, the scholars propose the creation of an International Data-Based Systems Agency (IDA) under the United Nations, modelled after the International Atomic Energy Agency (IAEA). The IDA would serve as a platform for global technical cooperation in digital transformation and DS among both state and non-state actors, promoting human rights, security and the peaceful use of DS. It would also act as a global regulatory body, overseeing a market authorisation process for DS.
“The IDA should work toward the sustainable, safe, and peaceful use of DS, contributing to international peace and security, the realisation of human rights, and the UN’s sustainable development goals,” Professor Kirchschläger explains, stressing that the global and inclusive approach of the proposed agency would help mitigate the risk of fragmentation in this critical area. Kirchschläger adds: “The IDA at the UN should be modelled on the IAEA as an ‘institution with teeth’.”
Additional information:
Policy Paper «Governing global existential AI Risks»
