Japanese group CODA takes OpenAI to task over Sora 2
Japanese content group CODA is requesting answers from OpenAI about Sora 2, its new text-to-video tool.
CODA says Sora 2 was trained using copyrighted Japanese works and is worried the tool might create videos too close to existing content—potentially breaking copyright laws.
They want OpenAI to be transparent about how it uses creators' work.
CODA argues that AI training without consent is illegal
CODA represents major names in Japan's creative industry and argues that, under Japanese law, using their content for AI training without permission isn't okay—even if OpenAI lets people "opt out" later.
They're asking OpenAI not to use members' works without clear consent and want honest responses if any copyright issues pop up.
This back-and-forth highlights a bigger debate
This back-and-forth highlights a bigger debate: as AI tools get smarter by learning from tons of online content, figuring out what's fair—and legal—for creators is only getting trickier.