RemoveAILabel

AI labelling rules by country

A creator-friendly overview of AI content labelling law in India, the European Union, China and the United States, and what it means for everyday posting.

Platform labels did not appear from nowhere. Governments have been writing disclosure rules for synthetic media, and platforms are building the systems to comply. This overview is written for creators, not lawyers, and laws change, so use it as a starting point.

Important. This page is general information, not legal advice. Dates and details were checked when the page was last reviewed and may have changed.

India

In February 2026 the government notified amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, in force from 20 February 2026. They introduce the term “synthetically generated information” for AI-made or AI-altered audio, images and video that appear real.

For creators: expect to be asked at upload whether your content is AI-generated, and answer accurately.

European Union

The EU AI Act includes transparency obligations for generative AI. Providers of AI systems must mark their output as artificially generated in a machine-readable way. People who deploy AI to create deepfakes must disclose that the content is artificial, with lighter treatment for evidently artistic or satirical work. These obligations were scheduled to start applying in August 2026, and the EU has been discussing adjustments to parts of the timetable, so check the current position.

For creators: if you publish realistic AI depictions of people, places or events to an EU audience, disclose it.

China

Measures on labelling AI-generated synthetic content took effect on 1 September 2025. They require both visible labels that people can see and implicit labels embedded in file metadata, and they place duties on services that generate and distribute content. Removing, altering or concealing these labels is prohibited.

United States

There is no single federal labelling law. Rules come from several directions:

For creators: political content, endorsements and advertising carry the most risk.

United Kingdom and elsewhere

The UK has relied mainly on online safety law, advertising standards and existing offences. Many other countries are drafting disclosure rules along similar lines to those above.

What stays the same wherever you are

  1. Using someone’s likeness or voice without consent can be unlawful with or without a label.
  2. Misleading buyers is unlawful with or without AI.
  3. Platform rules apply worldwide and are often stricter than local law.
  4. Honest disclosure is the lowest-risk position in every system described here.

Practical next steps: how to label AI content and when a label can and cannot be removed.

Questions people ask

Is it illegal to post AI content without a label?

It depends on the country and the content. Several jurisdictions now require AI-generated or deepfake content to be disclosed, and most place the main duties on platforms and AI providers. Deceptive use can also break existing fraud, defamation and consumer laws.

Do these laws apply to individual creators?

Some duties fall on creators directly, such as declaring AI content when a platform asks, or disclosing deepfakes under the EU rules. Others apply to the tools and platforms you use.

Is removing an AI label illegal?

In some places rules specifically target removing or altering AI labels and metadata, for example India’s 2026 rules for intermediaries and China’s labelling measures. Even where no specific rule exists, hiding AI origin in order to deceive can create liability.

Is this legal advice?

No. It is a general overview for creators. For decisions that carry legal risk, speak to a qualified lawyer in your country.