German Foreign Minister Johann Wadephul urged UN members on Saturday to develop shared safety rules for powerful AI models. He said providers should be required to test models before release and report serious safety flaws, backed by effective national and regional oversight.
Wadephul told the General Assembly that AI was already transforming societies and institutions needed to catch up. He called for greater capacity to understand and assess increasingly powerful systems, saying technological progress should serve human dignity.
Wadephul delivered the speech in New York on behalf of Chancellor Friedrich Merz and renewed his call for far-reaching UN reforms. The article says the UN's credibility crisis and failure to maintain peace and international security are weakening its authority to shape global rules.
