Vietnam AI Law Article 11: machine-readable marks for AI media

Vietnam's Law No. 134/2025/QH15, in force 1 March 2026, makes AI providers mark audio, image and video in machine-readable form. Sume docs show no such mark.

4 min readSume
All posts

Article 11 of Vietnam's Law No. 134/2025/QH15, effective 1 March 2026, requires providers of AI systems to mark audio, image and video content in a machine-readable form. The Government is to issue the detailed rules. Sume's public docs do not describe writing a machine-readable mark into generated files, so a provider-level mark is something I could not confirm from them.

Who owes what

The text is from an English translation on LuatVietnam, so rely on the official Vietnamese text for anything legal. The table separates the duties by who owes them.

Article 11 duties by role (read 2026-10-04)
RoleDuty in the translationSume docs say
ProviderMark audio, image and video in machine-readable formNo output mark documented
DeployerClear notice when content could confuse authenticityCaller's responsibility
DeployerEasily recognisable label on AI content simulating real people or real eventsCaller's responsibility
GovernmentIssue the detailed rulesNot yet read

What you do have

Sume's image API keeps the caller's metadata on the job. That is a useful internal record of which file was AI-made, but it is stored on the job and is not part of the file a viewer receives.

Timing and detail

The law takes effect on 1 March 2026. Article 11 leaves the detailed marking rules to the Government, so the exact format of a machine-readable mark was not in the text I read.

What I could not verify

Whether Sume counts as a provider, a deployer, or neither for a given customer depends on the facts and the Government's rules, and I did not verify that. Ask counsel before relying on any reading.

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume