Where Do Discussions on the AI Framework Act Stand Now?
Regulation-related talk keeps coming up in the news, but I’m not sure what will actually apply and from when.
I’m especially curious about the criteria for classifying high-risk AI. From a startup’s perspective, that one criterion can change the direction of a service. If anyone knowledgeable could break it down, I’d appreciate it.
7 answers
It passed the National Assembly late last year and goes into effect on January 22, 2026. Most of the 'regulations' you see in the news are referring to that, yeah.
The high-risk criteria are seriously ambiguous... In the end, won't we only know once the enforcement decree comes out?
As far as I know, the law itself only includes a rough definition of high-risk AI, and delegates the specific scope to a presidential decree. So I understand that the Ministry of Science and ICT is currently gathering industry feedback while drafting the enforcement decree. Employment, healthcare, finance, and public decision-making are the priority areas for discussion, and the key issue is how far to include generative models as well. If you're a startup, you can contact the Ministry of Science and ICT or the AI Legislation and Systems Division to get information on briefing session schedules.