The United States, China and the European Union signed the first report on the risks of AI

China, the United States, the European Union and around 20 other countries signed the Bletchley Declaration on Wednesday for the “safe” development of artificial intelligence (AI) in the UK. This happened during the first international summit on the rapid development of this technology.

“This historic announcement marks the start of a new global effort to strengthen public trust by ensuring artificial intelligence is safe,” British Prime Minister Rishi Sunak said on X (formerly Twitter).

The European Union and 28 countries gathered at Bletchley Park, north London, to “urgently understand and manage the potential risks” of AI, agreeing “through a new global initiative to ensure that AI is developed and used safely”. and responsible behavior.”

Addressing the growing potential of models like ChatGPT, the Bletchley Declaration “shows the world is coming together for the first time to recognize the problem and highlight its potential,” British Technology Secretary Michelle Donelan told AFP.

He added that the purpose of the meeting was “not to lay the groundwork for a global law, but to create a path forward”.

Two more international summits on AI will be held in South Korea in six months and in France in a year.

Political leaders, AI experts and tech giants met for two days in the UK’s bid to lead global cooperation in the field of AI.

Generative AIs, capable of generating text, sound or images in seconds at a simple request, have made exponential progress in recent years, and the next generations of these models will arrive this summer.

While they hold immense promise for medicine and education, the British government has warned that they can destabilize communities, enable weapons production and escape human control.

In an open letter published, several “founding fathers,” such as Joshua Bengio and Geoffrey Hinton, called for the “development and ratification of an international agreement on AI” to address the “catastrophic risks posed by advanced systems.” Humanity’.

The challenge is to define safeguards without stifling innovation from AI labs and tech giants. The EU and the US, unlike the UK, have opted for regulation.