Scope of Affected Models

The watermarking initiative specifically targets outputs generated by ChatGPT and Codex models to improve transparency for EU users.

Sources

Open the full topic