What Article 50 asks
- Providers of AI systems that generate images, audio, video or text must mark the output in a machine-readable way so it can be detected as AI-generated (Article 50(2)). Image generators do this with metadata such as C2PA Content Credentials and with watermarks.
- Deployers, the businesses that use those systems, must disclose deepfakes: image, audio or video that resembles real people, objects, places or events and could falsely appear authentic (Article 50(4)).
- Deployers must also disclose AI-generated text published to inform the public on matters of public interest, unless a person reviewed it and someone holds editorial responsibility for it (Article 50(4)).
- The rules apply from 2 August 2026. The Commission's code of practice on marking and labelling AI-generated content adds detail on how labels should look.
What PostBoost does
- One switch per post. Mark a post as AI-generated (
ai_generatedin the API) and PostBoost sets each network's own AI label where the network offers one through its API. - Automatic labels on our own AI output. Drafts written by PostBoost's AI features, and posts that use images generated by them, are labelled without you having to remember.
- Content Credentials kept when possible. When an image arrives with C2PA Content Credentials and does not need resizing or conversion, we store and publish the original file untouched. If an image has to be resized or converted, its credentials cannot survive, and we record that.
- AI usage records. AI generations are logged per workspace, so you can show which content came from AI.
Labels by network
| Network | What PostBoost sets |
|---|---|
| TikTok | The "AI-generated content" label, on every TikTok account of the post. |
| Instagram, YouTube | Not yet available through PostBoost. The API answers X-AI-Disclosure: partial so you know to add the label in the app or YouTube Studio, or note it in the caption. |
| Facebook, LinkedIn | No API switch. These networks read Content Credentials in images when they are present, which is why we keep the original file when we can. |
| X, Threads, Bluesky, Mastodon, Pinterest, Google Business Profile, Telegram, Discord, Slack | No AI label in the API. Add a note such as "Made with AI" to the caption when a disclosure is required. |
What stays with you
- Deciding what counts as AI-generated or AI-manipulated content in your posts, including content made with tools outside PostBoost.
- Reviewing AI-written text before it is published, and taking editorial responsibility for it.
- Disclosing deepfakes clearly in the post itself where the network offers no label.
- Keeping Content Credentials intact in your own workflow: editing an image in a tool that strips metadata removes them before PostBoost ever sees the file.
For developers
Send "ai_generated": true when you create or update a post through the API or MCP. The AI disclosure guide lists the exact behaviour per network.
This page describes what PostBoost does. It is not legal advice and does not, on its own, make your content compliant. Last reviewed October 11, 2026. Questions: [email protected].