Britain And Allied Countries Agree To Test Artificial Intelligence Models Before Release

JAKARTA - British Prime Minister Rishi Sunak on Thursday 2 November announced that the governments that exist and artificial intelligence companies have reached an agreement to pilot a new AI model before its release.

Sunak hosted the AI Safety Summit for two days where on Wednesday 1 November, he managed to gain support from China for international efforts to manage the risks of the technology.

On Thursday, Sunak announced a further "historic" agreement with "governments that trust and AI companies."

"We will work together in testing the security of the new AI model before it is released," Sunak said at a press conference at the peak of the meeting.

The list of countries that have joined the security testing collaboration does not include China, whose representatives were not invited on the second day of talks.

Representatives from the United States, European Union, France, Germany, Italy, Japan, South Korea, Singapore, Canada, and Australia were involved in the talks.

"It is crucial that future AIs are safely developed and that the potential risks of new models are rigorously evaluated before and after being implemented, including by evaluating possible dangerous capabilities," read a British statement as chairman on the summary of participants' discussions.

Asked why China was not invited to discuss AI model testing, Sunak said it was appropriate to invite China to part of the meeting, in line with the UK's policy of engaging with China as far as possible but also protecting itself from risk.

"We achieved the results we wanted, namely they are here, involved... and they have signed the Bletchley Park communique," Sunak was quoted as saying by VOI from Reuters.