OpenAI and Microsoft complain that DeepSeek may be trained on their data, “obtained in an unauthorized manner” via the API.
Now they know how publishers feel, I guess 😉
https://www.reuters.com/technology/microsoft-probing-if-deepseek-linked-group-improperly-obtained-openai-data-2025-01-29/
Now they know how publishers feel, I guess 😉
https://www.reuters.com/technology/microsoft-probing-if-deepseek-linked-group-improperly-obtained-openai-data-2025-01-29/
Comments
@proeschold.bsky.social asked DeepSeek yesterday and noone seems to hide the fact:
https://bsky.app/profile/proeschold.bsky.social/post/3lgs53ron5s2u
Det er ikke usandsynligt at den er trænet på eksisterende modellers output - de er så imod deres TOS, men ikke noget de nemt kan gøre noget ved.
https://bsky.app/profile/troelsbj.bsky.social/post/3lgujl5owvs2z