VIFLY

The VIFLY guide France · European Union

Each platform has rules. Your audience keeps its rights.

The same content may need different settings on different platforms. An app label and a legally required disclosure address separate requirements.

Sources checked on 23 October 2026

CREATE WITH CONFIDENCE
01 / THE STARTING POINTA tool. A framework.I check what I use.
02 / THE RIGHT HABITRights respected.People, works, data.
03 / THE RIGHT INFORMATIONAn informed audience.Clear from publication.
Clarity first
Practical answersOfficial sourcesDiagrams & examplesWithout unnecessary jargon

The short answer

AI on social media

Disclose synthetic or altered content when the platform requires it, then separately check applicable French and European rules. An automatic label proves neither a person's consent nor the accuracy of the content.

LawPlatformGood practicePause and check

These guideposts distinguish the status of a rule; 'Pause and check' calls for review before acting.

The right order before posting

  1. The content

    Realistic? Altered? Sensitive?

  2. The platform

    Enable the required disclosure.

  3. The rights

    Check the law and permissions.

The platform label does not replace other checks.

From rule to practice

What does that look like in practice?

Platform

A realistic fake flood

The video looks like authentic local news.

Enable the required disclosure and explain that the scene is fictional; also check Article 50 and the risk of misleading the public.

Good practice

A caption with spelling corrections

AI corrects grammar without changing the meaning.

Do not turn this into a universal labelling duty. Check the final content and your platform's rule.

Pause and check

A very convincing fake client

An avatar claims a result no real client has achieved.

Do not publish this fake testimonial. An AI label does not make the claim truthful.

YouTube: realism and music in the foreground

YouTube's rule covers realistic content that is meaningfully altered or generated: making someone say something they did not say, altering a real event, or creating a realistic scene that never happened. Its help page also includes AI-generated music when it is a main element of the content.

In YouTube Studio, complete the AI use setting when uploading. Script assistance, captions and some minor edits are examples that do not require this disclosure. Check the official examples for your use case: the policy also discusses certain dubbing uses of your own voice.

YouTube states that disclosure alone does not limit audience or monetisation. That does not guarantee eligibility: other rules, including those on authenticity and misleading content, still apply.

TikTok: disclose realism and check the person

TikTok requires labels for realistic AI-generated images, audio and video. It also encourages labels for fully generated or substantially edited content. Native effects may trigger an automatic label.

Some content remains prohibited even with a label. TikTok rules restrict, among other things, synthetic depictions of minors and use of adult private figures' likeness without permission. Read the full policy before imitating or creating an avatar of a real person.

Instagram and Facebook: AI information and context

Meta uses technical signals and user disclosures to display AI information. Presentation may vary between generated and merely edited content. For photorealistic video or realistic-sounding audio digitally created or altered, use the platform's disclosure tool.

Keep a clear explanation in the content when context could mislead. Rules on impersonation, harassment, advertising and third-party rights continue to apply, with or without a label.

LinkedIn and X: do not present fiction as evidence

LinkedIn recommends disclosing substantial AI creation or editing when it is not obvious. That recommendation is distinct from a general legal duty. Community rules prohibit false or misleading content: a fake client testimonial or qualification remains a problem.

X restricts synthetic, manipulated or out-of-context media likely to cause widespread confusion on public issues, affect public safety or cause serious harm. This does not imply a uniform hashtag duty for every AI-assisted text. Also check the rules on non-consensual intimate media.

Before taking action

Your essential checks.

A preparation aid, with no automatic compliance verdict.

Getting to the point

Questions people often ask.

Does an AI label automatically remove YouTube monetisation?

YouTube states that disclosure alone does not reduce audience or remove monetisation. Other eligibility criteria and content rules still apply.

Is an automatic label always enough?

No. Check its presence and clarity, the platform's exact requirements and applicable legal duties. It is not permission to use a person or a work.

Sources and method

The sources behind the answers.

Editorial team: VIFLY. Documentary checks on 23 October 2026. This guide is an educational summary without individual legal validation. The links let you consult the text and its conditions.

Understand, then practise

Clear rules. Useful applications.

Explore AI courses and VIFLY resources to build a more considered practice.