MUSEBLOSSOM

C2PA manifests, invisible watermarks and fingerprints answer different questions. What each proves, where it fails, and which one can trace a leak.

Six questions to ask before buying an AI music detection tool, beyond the accuracy number.

How streaming services separate an AI detection flag from enforcement, what went wrong in two 2026 false-positive cases, and what a fair review process needs.

How YouTube Music decides which songs get an AI label: partner declarations plus YouTube's own signals.

Two AI detection figures, 99.8% and 96%, show why accuracy claims can't be ranked side by side, and the four questions to ask first.

AI watermark detectors mostly read their own company's marks. What SynthID Detector and OpenAI's textGrain detector check, and how C2PA Content Credentials differ.

MuseBlossom and Koryo Credit Information co-develop invisible watermarking with Image Defence to trace leaks from phone photos of work screens.

The European Commission opened a targeted copyright consultation on September 29, 2026. It covers AI use of protected works, live-event piracy, remuneration for non-EU recordings and scientific research. Non-EU organisations can respond until November 3.

85.6% of YouTube, Instagram and TikTok users in Korea say AI content should be labeled. Most labels come from file metadata, and that metadata is easy to lose.

An AI-composed track entered Melon's Top 100. The human lyricist received zero royalties — only neighboring rights. This exposes Korea's attribution gap in AI music copyright.

Copyright rules for AI-assisted music differ by country. A look at Korea's KOMCA reversal, the Randy Travis case, and what Suno, Udio, and Spotify's terms actually allow.

A photo or a piece of music travels far enough on the internet and, at some point, nobody can say who made it — or whether anyone made it at all. C2PA is the standard built to address that. Here is what it is and why it matters, with diagrams and the vocabulary you'll run into.

Two techniques that keep showing up in the same sentence in every Article 50(2) compliance guide — and why neither one alone gets you there.

The transparency obligations for AI systems and AI-generated content, article by article — and who actually has to comply.