2024-03-22 03:59:21
The United Nations General Assembly has approved its first global resolution on artificial intelligence (AI), which aims to ensure the technology is “safe, secure and reliable,” Reuters and the Associated Press reported.
The resolution aims to promote the protection of personal data, the monitoring of AI for risks and the protection of human rights, UN officials said.
The non-binding document, proposed by the US and submitted jointly with 122 other countries, was adopted by consensus without a vote – meaning it has the support of all 193 member countries. Negotiations on it were conducted for 3 months, notes Reuters.
“Today we are at a crossroads. Artificial intelligence poses universal existential challenges,” said US Permanent Representative to the UN Linda Thomas-Greenfield, referring in particular to the “deep fakes” that risk “undermining the integrity of political debate in a year when more from half the world will choose its leaders”.
“However, artificial intelligence also provides significant universal opportunities to accelerate the fight against poverty, save human lives and protect our planet,” she stressed, recalling the benefits that she believes are already being seen in medical diagnostics and agriculture.
The resolution also recommends strengthening policies for the protection of personal data, officials of the organization told the media before the adoption of the resolution.
US National Security Adviser Jake Sullivan said earlier this month that the resolution’s passage would be a “historic step forward” in promoting the safe use of AI. The resolution “will represent global support for a core set of principles for the development and use of AI and will chart a path to use AI systems for good while managing risks,” he stressed in a statement to the AP.
“We are navigating turbulent waters with rapidly changing technology, which means it is more important than ever to be guided by the light of our values,” said one senior administration official, describing the resolution as “a first-of-its-kind document on a true global consensus on AI”.
The resolution is the latest in a series of initiatives, mostly non-binding, by the world’s governments trying to guide the development of AI amid concerns that it could be used to disrupt democratic processes, facilitate fraud or lead to dramatic loss of jobs, among other harms.
“The improper or malicious design, development, implementation and use of artificial intelligence systems creates risks that could undermine the protection, promotion and exercise of human rights and fundamental freedoms,” the measure said. Asked whether the negotiators faced resistance from Russia or China – the UN member states that also voted in favor of the document – the officials acknowledged that there had been “very heated discussions.” But active contacts have been made “with China, Russia, Cuba and other countries that often disagree with us on some issues,” they added.
“We believe the resolution strikes an appropriate balance between furthering development while continuing to protect human rights,” said one of the officials, speaking on condition of anonymity.
In November, the US, Britain and more than a dozen other countries unveiled the first detailed international agreement aimed at preventing unscrupulous actors from using AI, insisting that companies create AI systems that are “secure by design”.
Europe has overtaken the US, with EU lawmakers this month adopting an interim agreement to oversee the technology, moving closer to adopting the world’s first artificial intelligence rules. The Biden administration has been pushing lawmakers to regulate AI, but the polarized US Congress hasn’t made much progress.
The White House sought to reduce AI-related risks to consumers, workers and minorities while strengthening national security with a new executive order passed in October.
Like other countries around the world, Russia and China are eagerly exploring the use of AI tools for various purposes. Last month, the company “Microsoft” announced that it had detected hackers from both countries who were using the company-supported “OpenAI” project to improve their hacking skills, writes BTA.