Singapore PM Lawrence Wong proposes AI incident reporting pilot modelled on aviation safety

Prime Minister Lawrence Wong has proposed an artificial intelligence incident reporting pilot modelled on international aviation safety, where countries voluntarily share lessons from AI failures and near misses. Speaking at the Virtual Leaders' Meeting on Frontier AI Safety on 28 Oct, he said inter

SINGAPORE: Prime Minister Lawrence Wong has proposed that interested parties and countries launch an artificial intelligence (AI) incident reporting pilot, drawing lessons from international aviation where incidents and near misses are reported and studied. Speaking at the Virtual Leaders' Meeting on Frontier AI Safety on Wednesday (28 Oct), Mr Wong suggested such a pilot could be a first step in AI governance.

“We can start by agreeing on a common set of incident categories and information to be shared,” he said. “We can explore how participating countries can collect and exchange this information securely. And this will allow us to test what works, identify the practical difficulties, and build confidence before considering a wider framework.”

Rather than seek to establish an “elaborate new international architecture” from the outset, Mr Wong said countries could organise cooperation around specific workstreams, with smaller cross-regional groups of interested countries taking the lead. “The participation and adoption can remain voluntary, so countries can contribute according to our interests and capabilities,” he added.

The meeting, co-hosted by Norwegian Prime Minister Jonas Gahr Støre and Finnish President Alexander Stubb, involved 22 countries and the European Commission. It was a follow-up to the joint declaration for the control of frontier AI models at the United Nations General Assembly in September, which Mr Wong and more than 20 world leaders signed, calling for stronger safeguards for frontier AI.

Concerns over frontier AI safety have recently made headlines. In September, a rogue AI agent from OpenAI breached an Australian national healthcare database, prompting Australian leaders to consider possible “law enforcement and legislative responses”. In July, about 700 AI agents created by OpenAI hacked the open-source platform Hugging Face. OpenAI said in October it had resumed training of its most advanced AI models after tightening security and barring internet access.

Mr Wong said Singapore stands ready to do its part in advancing safe and responsible AI use, and that AI safeguards should build on existing efforts rather than duplicate them. Singapore convenes an International Scientific Exchange on AI Safety that has produced the Singapore Consensus on Global AI Safety Research Priorities. “We can also draw on the work of the Independent Scientific Panel on AI, which supports the UN Global Dialogue on AI Governance,” he said.

Highlighting AI incident reporting as one specific area for practical cooperation, Mr Wong said lessons should not be confined to one country or one company when serious AI incidents happen. “We should be able to learn from what went wrong, identify the emerging risks, and take steps to prevent similar incidents happening elsewhere,” he said. He noted that current knowledge about frontier model safety mostly comes from developers, but that governments, businesses and other organisations using these models also encounter serious failures or near misses that can offer valuable lessons, even if no actual harm has occurred.