Dozens of countries call for ‘safe, reliable and robust AI’

Dozens of countries are calling for "safe, reliable and robust AI". Keystone-SDA

Dozens of countries, including the United States, China and Switzerland, have called for safe, trustworthy and robust artificial intelligence (AI).

+Get the most important news from Switzerland in your inbox

They signed a joint declaration published on Saturday at the end of the AI summit in New Delhi.

“Promoting safe, trustworthy and robust AI is essential for building trust and maximising social and economic benefits,” said the text signed by 86 countries and two international organisations.

The text does not contain any concrete commitments, but it does highlight several voluntary and non-binding initiatives, particularly with regard to pooling AI research capabilities at the international level.

+ AI-driven discrimination: protection in Switzerland is weak

“We believe that the potential of AI can be fully realised only when its benefits are shared by all of humanity,” states the declaration published at the end of the five-day summit. It describes the advent of AI as “an inflection point in the trajectory of technological evolution”.

“The commitments made are not completely insignificant; the main thing is that there are commitments, quite simply,” Stuart Russell, a leading computer scientist, told AFP.

“I hope that all countries will be able to build on these agreements […] to develop legally binding commitments to protect their populations, so that the development and deployment of AI can continue without posing unacceptable risks,” Russell continued.

Dozens of delegations

Dozens of delegations from around the world gathered in the Indian capital this week to discuss, among other things, how to regulate this rapidly growing technology.

Indian Prime Minister Narendra Modi, Swiss President Guy Parmelin, French President Emmanuel Macron, UN Secretary-General António Guterres and leading tech executives, including Sam Altman of Open AI, spoke at the summit, which opened on Thursday.

Among the hot topics were the societal benefits of multilingual AI translation, threats to employment, and the issue of data centre energy consumption.

Two divergent visions

During the debates, two different approaches to managing the development of AI were put forward. Guterres confirmed the launch of a scientific commission aimed at making “human control” of AI “a technical reality”.

Pointing out that the United States and China currently control the vast majority of AI research and investment, Parmelin called for “a monopoly or duopoly to be avoided”. Some large countries have taken the lead. It is essential to restore the balance, he said.

Other countries must organise themselves and create global governance, he stressed, citing India, South Korea, France and Switzerland. On Friday, the United States “totally” rejected any global governance of artificial intelligence, through the voice of Michael Kratsios, White House advisor on science and technology.

At the 2025 edition in Paris, the United States did not sign the declaration. Altman said on Thursday that this rapidly growing technology urgently needs regulation. The next summit will take place next year in Geneva, Parmelin announced.

Adapted from French by AI/ts

How we work 

We select the most relevant news for an international audience and use automatic translation tools to translate them into English. A journalist then reviews the translation for clarity and accuracy before publication.  

Providing you with automatically translated news gives us the time to write more in-depth articles. The news stories we select have been written and carefully fact-checked by an external editorial team from news agencies such as Bloomberg or Keystone.

If you have any questions about how we work, write to us at [e-mail]