Vietnam Shares Its 46-System High-Risk AI List With Australian Officials at Quang Ninh Forum

Vietnam’s list of high-risk artificial intelligence systems now covers 46 systems, officials told their Australian counterparts at the Vietnam–Australia digital cooperation forum in Quang Ninh on October 28, VnExpress reported. The forum was part of Vietnam International Digital Week 2026, and much of the discussion focused on how a risk-based approach to AI governance works in practice.

Tom Marwick of Australia’s Digital Transformation Agency asked how Vietnam combines its risk classification with impact assessments while keeping common standards and some flexibility. Chu Thi Tham, who heads digital transformation at the National Institute of Digital Technology and Digital Transformation under the Ministry of Science and Technology, said Vietnam defines high-risk AI by its potential to significantly affect health and life, the rights of organisations and individuals, the public interest and national security.

Some systems on the list must be assessed by an independent third party before they operate, while others can be self-assessed by the provider or checked by a hired assessor before going to market. Sector ministries propose entries for their own fields, and the Ministry of Science and Technology reviews them before the government issues the list. For the public sector, Tham said, humans must keep the final decision and legal responsibility.

What it means in practice

Hoang Huu Hanh, deputy director of the National Digital Transformation Agency, said the list cannot be fixed. A system rated low risk while a person controls it directly can become high risk as automation increases and it begins to affect safety or national security. He pointed to controlled testing as a way to check new systems before broad rollout, a role Vietnam’s generative AI sandbox, which does not grant deployment rights, already plays.

Lucy Poole of the Digital Transformation Agency described Australia’s three-part toolkit: an AI technical standard, an impact assessment tool built on eight ethics principles, and an advisory review board whose recommendations are not binding. For companies in Vietnam, the practical question is which sector lists apply to them, as banks have learned since the State Bank of Vietnam set risk tiers and human review rules for banking AI.

Hanh said Vietnam has shaped its legal framework and the next challenge is putting rules into practice, which needs resources and experience. What to watch next is how often the list is updated, which assessors are accredited for third-party checks, and what follows from a week that put AI governance at the top of the agenda in Quang Ninh.