Insights from UN Security Council’s AI meeting, key takeaways

The UN Security Council convened for its first-ever debate on AI, delving into the technology’s potential impact on global peace and security. The meeting, chaired by UK’s Foreign Secretary James Cleverly, witnessed testimonies from experts, shedding light on the developments in AI and the risks it poses. The officials called for a sense of urgency and a global perspective in tackling AI’s challenges. As AI innovation shows no signs of slowing down, the race for AI dominance and technological supremacy heightens the urgency for comprehensive regulations and governance.

The council received insights from experts Jack Clark, co-founder of Anthropic, and Professor Zeng Yi, co-director of the China-UK Research Center for AI Ethics and Governance. 

UN Secretary General

The call for action to address AI’s risks for the sake of present and future generations

UN Secretary-General António Guterres also delivered a speech where he once again called for the establishment of a global agency to oversee AI, highlighting the urgent need for action to address potential catastrophic risks associated with AI. Guterres stressed the dual nature of AI, emphasising its positive impact on peacekeeping, humanitarian efforts, and sustainable development while acknowledging its potential for malicious use in terrorism, cyberattacks, and disinformation. Furthermore, he said that the UN must formulate a legally binding agreement before 2026 to prohibit the use of AI in automated weapons of war.

Secretary-General’s remarks to the Security Council on AI

‘Generative AI has enormous potential for good and evil at scale. Its creators themselves have warned that much bigger, potentially catastrophic and existential risks lie ahead. Without action to address these risks, we are derelict in our responsibilities to present and future generations.’

Experts insights

Jack Clark urged governments to play a significant role in AI development, stating that leaving it solely to the private sector is inadequate, adding the importance of understanding AI to avoid unanticipated misuses. Clark also highlighted the challenge of chaotic behaviour in AI systems, urging the audience to recognise AI as a form of human labour that gains power and accessibility over time.

Professor Zeng Yi called on the UN to establish a framework for AI development and governance, ensuring global peace and security. In his opinion, UN must play a central role in this process. Finally, Yi cautioned that AI could pose a risk of human extinction if measures to protect against exploiting human weaknesses are not implemented.

United Kingdom’s vision of governing AI

UK Foreign Secretary James Cleverly presented the UK vision of governing AI, which is based on four irreducible principles:

Foreign Secretary also added that the UK’s strategy leverages established multilateral efforts like the AI for Good Summit and the work of UNESCO, the OECD, and the G20; it focuses on collaboration with key institutions and initiatives such as the Global Partnership for AI, the G7’s Hiroshima Process, the Council of Europe, and ITU; and aims to achieve close cooperation between pioneering AI companies and the UK to ensure the benefits are maximised, and the risks to humanity are mitigated.

United States, China, and Russian perspectives

Chinese Ambassador Zhang Jun acknowledged the dual nature of AI and emphasized the importance of establishing standards and guidelines for responsible development. China supported the UN taking a central role in setting guiding principles for AI. They also stressed the significance of enhancing global understanding of AI ethics, safety, reliability, and control.

The US Deputy Ambassador, Jeffrey DeLaurentis, expressed concerns about AI misuse, particularly censorship and repression. The US highlighted the need for international collaboration to address human rights risks and ensure global peace and security.

In contrast, Russian Ambassador Dmitry Polyanskiy questioned whether discussing AI was within the council’s purview. Russia suggested a scientific and expertise-based discussion should take place over several years, believing such conversations were already happening on specialised platforms.

AI-generated report

Here you can find an AI-generated report from the UN Security Council discussion on AI hosted by the United Kingdom on 18th July 2023. This just-in-time reporting is provided by DiploGPT, a sophisticated, domain-specific artificial intelligence solution designed to exploit the capabilities of advanced natural language processing technologies.

DiploGPT combines state-of-the-art speech-to-text, information retrieval, text generation, and text-to-voice models – both proprietary and open-sourced – to create a specialized, high-performance tool for diplomatic use cases. Utilizing cutting-edge fine-tuning methodologies, DiploGPT is refined by incorporating the knowledge of subject-matter experts in diplomacy and linguistics, which enables the adaptation of large language models for optimal efficacy in diplomatic scenarios. To learn more about this AI approach, please visit Diplo’s AI page. You can also consult updates from Diplo’s experts.