The United Nations General Assembly has unanimously consented to adopt what some call the first global resolution on AI. The agreement is nonbinding, but it gives governments a shared international text on safe, secure and trustworthy artificial intelligence systems.
The resolution grew out of a proposal by the United States and received backing from China and 121 other countries. Its central goals are to protect personal data, strengthen privacy policies, monitor AI for potential risks, and uphold human rights.
What the UN resolution is trying to do
The resolution is titled Seizing the opportunities of safe, secure and trustworthy artificial intelligence systems for sustainable development. That title captures the balance the text is trying to strike: encourage the use of AI for development while recognizing that irresponsible or malicious uses of AI systems can put human rights and fundamental freedoms at risk.
The agreement focuses on several broad priorities:
- Protection of personal data.
- Stronger privacy policies.
- Close monitoring of AI systems for potential risks.
- Support for human rights and fundamental freedoms.
- A balance between AI development and safeguards.
The resolution does not create binding enforcement. That matters because it limits what the agreement can compel countries or companies to do. Still, the adoption gives the UN a common reference point for discussing AI governance at a global level.
Why consensus matters
In the UN, adoption by consensus means the text moves forward without a vote because all members agree to adopt it. The UN explains that this does not mean every Member State agrees with every part of a draft document.
“Consensus is reached when all Member States agree on a text, but it does not mean that they all agree on every element of a draft document,”
The same UN FAQ says Member States can adopt a draft resolution without a vote while still holding reservations about parts of the text. That distinction is important here because AI policy has become a difficult international issue, and the resolution followed three months of negotiation.
US officials acknowledged that there were “lots of heated conversations” during the process, according to Reuters. Resistance from nations such as Russia and China had been anticipated, yet the negotiations still produced a draft that could be accepted by all UN countries.
The industry response
Because the agreement is nonbinding, it appears to be broadly popular in the AI industry. Microsoft Vice Chair and President Brad Smith welcomed the adoption on X.
“We fully support the @UN’s adoption of the comprehensive AI resolution. The consensus reached today marks a critical step towards establishing international guardrails for the ethical and sustainable development of AI, ensuring this technology serves the needs of everyone.”
That response reflects one of the resolution’s main themes: international guardrails for AI without the immediate force of binding regulation. The text gives governments and companies a broad framework for discussing AI safety, privacy, development, and rights, while leaving future action to other policy processes.
A senior US administration official also framed the moment around shared values during rapid technological change.
“We’re sailing in choppy waters with the fa st-changing technology, which means that it’s more important than ever to steer by the light of our values,”
The same official described the resolution as a “first-ever truly global consensus document on AI.”
How it fits into global AI governance
The new UN agreement may be the first “global” AI agreement in the sense that every UN country participated. It is not, however, the first multi-state international AI agreement.
The source article notes that the Bletchley Declaration, signed in November by the 28 nations attending the UK’s first AI Summit, appears to be the earlier multi-state agreement. Also in November, the US, Britain, and other nations unveiled an agreement focused on creating AI systems that are “secure by design” to protect against misuse by rogue actors.
Governments are moving at different speeds. Europe is slowly moving forward with provisional agreements to regulate AI and is close to implementing the world’s first comprehensive AI regulations. The US government, meanwhile, still lacks consensus on legislative action related to AI regulation, while the Biden administration has advocated measures to reduce AI risks and enhance national security.
The stakes behind the agreement
The resolution arrives after the launch of ChatGPT and GPT-4 and after the enormous hype raised by certain members of the tech industry in a public worldwide campaign last year. That attention has made AI development a central issue for governments, companies, and critics.
Critics fear AI could undermine democratic processes, amplify fraudulent activities, or contribute to significant job displacement, among other concerns. The UN resolution does not resolve those issues by itself. Its value is more basic: it records a global willingness to treat AI as a technology that must be developed with attention to privacy, personal data, human rights, and risk.
For now, the result is a shared but nonbinding framework. It signals cooperation across countries that often disagree, while leaving the harder questions of enforcement, regulation, and implementation for the next stage of AI governance.