Guides de nettoyage des métadonnées par plateforme
Découvrez comment chaque réseau social et place de marché inspecte les photos, quelles métadonnées déclenchent des badges automatisés et comment préparer vos images sans perte de qualité.
Instagram flags images with 'AI info' badges by reading embedded C2PA Content Credentials and IPTC DigitalSourceType tags. If you edited a camera shot in Photoshop or Lightroom with minor generative tools, Instagram may incorrectly tag your authentic photo as AI.
Voir le guide Instagram →Meta systems apply 'AI info' labels across Facebook feeds based on metadata standards defined by the Coalition for Content Provenance and Authenticity (C2PA).
Voir le guide Facebook →Threads (EN)
Threads shares Meta's automated provenance detection pipeline, checking uploaded media for C2PA manifests and algorithmic media tags.
Voir le guide Threads →TikTok
TikTok actively checks images for C2PA Content Credentials and applies an automated 'AI-generated' label to posts detected with provenance markers.
Voir le guide TikTok →Pinterest (EN)
Pinterest indexes metadata and categorizes pins based on embedded creator information, software strings, and prompt parameters.
Voir le guide Pinterest →Etsy (EN)
Selling handmade goods, vintage items, and creative prints on Etsy requires privacy and compliance with platform creation policies.
Voir le guide Etsy →LinkedIn (EN)
LinkedIn professional profiles and posts often feature headshots and design assets containing hidden camera serials, GPS data, and tool markers.
Voir le guide LinkedIn →X (formerly Twitter) (EN)
While X re-encodes most images to web-optimized formats, original EXIF and metadata can leak sensitive personal information during direct uploads or API interactions.
Voir le guide X (formerly Twitter) →Reddit (EN)
Many Reddit communities (e.g. photography, art, and privacy subreddits) inspect image metadata or have strict rules regarding AI tags.
Voir le guide Reddit →