Singapore Prime Minister Lawrence Wong : AI incident reporting is one way for countries to strengthen safety cooperation

SINGAPORE: After Singapore joined the list of signatories calling for the control of frontier AI models last month, Prime Minister Lawrence Wong provided an update to the initiative headed by Norwegian Prime Minister Jonas Gahr Støre and Finnish President Alexander Stubb.

In a social media post on Wednesday (Oct 28), PM Wong said that he had joined a virtual meeting that the two European leaders hosted, where different leaders talked about how to take the initiative forward.

The Prime Minister explained that the discussion centred on practical ways to strengthen cooperation for AI safety, and added that one such area is AI incident reporting, as this allows for learning how emerging risks may be identified and thwarted.

He added that this practice is done in international aviation, as near-misses are studied in order to improve safety.

“We should build a similar culture of learning and prevention for AI, sharing lessons from governments, businesses and other organisations using frontier AI models, including serious failures and near misses even if no actual harm has occurred,” wrote PM Wong in his post.

Countries that are interested in AI incident reporting could start with a pilot programme that would develop and test practical approaches for a secure sharing of such information.

“This will help us learn what works and build confidence before moving towards a wider framework,” he added.

The statement calling for controlling AI frontier models was originally signed by representatives from 21 countries, including a number of European leaders, such as those from Germany, Denmark, Spain, and the Netherlands; the President of the European Commission; and also by the Prime Ministers of Australia and Canada; the Presidents of Kenya and South Africa; and leaders from Türkiye, Bahrain, and the United Arab Emirates.

Shortly afterwards, nine other countries followed suit and signed the call, which can be found here. 

Conspicuously absent from the signatories are the leaders of the two global superpowers, the United States and China.

The call noted warnings from scientists and the executives of tech companies about AI developing too quickly to effectively manage risks, and called for immediate action from the industry, governments, and society.

The statement asked for the following:

1. Companies to develop transparent safety protocols, including mandatory pre-deployment testing and independent evaluation, with qualified evaluators granted sufficient access to assess risks. 

2. Governments and regional organisations to further develop and coordinate common standards, strengthen transparency — including shared reporting of serious safety incidents — and ensure that countries across all regions have access to scientific capacity, expertise and trusted evaluation. /TISG

Read related: Trump dismisses AI safety concerns as ‘hoax’ as tech leaders call for slower development