105
Chinese labs DeepSeek and Moonshot were quietly relaying customer prompts to Claude through fraudulent accounts. Users thought they were using Chinese AI, but were actually getting answers from Claude (www.anthropic.com)
submitted 1 month ago by to c/technology@lemmy.world
16 comments fedilink hide all child comments
 

cross-posted from: https://piefed.world/c/tech/p/1393607/chinese-labs-deepseek-and-moonshot-were-quietly-relaying-customer-prompts-to-claude-thro

all 16 comments

sorted by: hot top controversial new old
[–] 84 points 1 month ago

  • permalink
  • fedilink
  • source
    •
    [–] 76 points 1 month ago

    is anthropic.com a reliable source for an information that benefits anthropic's image?

  • permalink
  • fedilink
  • source
    •
    [–] 63 points 1 month ago (1 child)

    I don't believe a word they say.

    They are lying for their stupid IPO

  • permalink
  • fedilink
  • source
  • hideshow 1 child comment
    •
  • [–] 2 points 1 month ago

    I love open Chinese llms but this seems 100% plausible and and I think ur wring. This is in line with how Chinese companies need to work around the system since America tries to handicap them. This is actually a very smart way to get high quality trianjng data for their future models. Unfortunate about leaking private customer data to a us company though.

    Also the article and the evidence presented by Antropic all seems very plausible.

  • permalink
  • fedilink
  • source
  • parent
    • [–] 47 points 1 month ago (1 child)

    Proof: trust me bro.

  • permalink
  • fedilink
  • source
  • hideshow 1 child comment
    •
    • [–] 45 points 1 month ago

    The onion is nothing compared to real headlines

  • permalink
  • fedilink
  • source
    •
    [–] 13 points 1 month ago (3 children)

    ummm, the only times I ever used deepseek was fully offline on my hardware, I call bullshit

  • permalink
  • fedilink
  • source
  • hideshow 3 child comments
    •
  • [+] -7 points 1 month ago (2 children)

    You almost certainly weren't using a Deepseek model offline (unless you have 100s of GB of RAM/VRAM). You might have been using a small model distilled on Deepseek responses, but the smallest model made by Deepseek is about a 300B param model.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 13 points 1 month ago (1 child)

    the smallest model made by Deepseek is about a 300B param model.

    This is not correct. DeepSeek has plenty of smaller models. I just had a scroll through their huggingface, and picked out a few: DeepSeek-V2-Lite (16B), deepseek-moe-16b-chat, deepseek-llm-7b-chat, but there's quite a few more too.

    Though DeepSeek R1 is their flagship model, which is the one people usually think of, and that's a 684B param model.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 6 points 1 month ago* (last edited 1 month ago)

    huh fair enough, I hadn't seen those before, they're from before deepseek made a name for themselves (which as you say was with R1). They are very out of date though, as is R1 (from the end of 2024 and start of 2025 respectively). the current set of models from Deepseek come in 291B and 1.6T varieties and the previous version only was released in 685B

    You'd almost certainly be better off upgrading to qwen or Gemma4 for a local model of that kind of size.

  • permalink
  • fedilink
  • source
  • parent
    • [–] 10 points 1 month ago*

    Alright, so what? Also this (even if it's true and not another clown show by claude) won't affect models that aren't hosted by moonshot or deepseek.

  • permalink
  • fedilink
  • source
    •
    [–] 10 points 1 month ago

    What this really boils down to is that Anthropic and OpenAI hope to be the only ones that use someone else's copyrighted information without any legal or financial complications.

  • permalink
  • fedilink
  • source
    •
    [–] 5 points 1 month ago (1 child)

    You're telling me that is routing requests through a model 70x the price?

  • permalink
  • fedilink
  • source
  • hideshow 1 child comment
    •
    • [–] -5 points 1 month ago

    Yeah that's pretty crazy but this bit is crazier - https://www.anthropic.com/threat-intelligence-report-september-2026#cyber-operations-sep-26

  • permalink
  • fedilink
  • source
    •