Anthropic says Alibaba illicitly extracted Claude AI model capabilities • Hacker News Board

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#471

Post by dosisking » Thu, Jun 25, 2026, 6:56 AM UTC

user_7832 wrote: post #357

petterroea wrote: post #306

Earlier quoted context omitted.

China aren't offering a cheaper solution. They are subsidizing an existing one (which is already subsidized) in order to gain foothold. The difference is that in the US subsidies come from VC, while OP implies subsidies come from the AI labs that buy the training data (which may as well also be VC backed, so just one extra hop). This isn't "the market working as intended", this is an exhaustion fight to the bottom wh…

> China aren't offering a cheaper solution. They are subsidizing an existing one (which is already subsidized) in order to gain foothold. In my economics classes, we were told that (in a "free market" argument) the best thing to do if a subsidy is making something you want cheaper is to use it. You're getting your thing, and at a reduced cost. (I'm not really replying to you per se, I'm curious how "free market" folk…

Most everything they teach in economics class is wrong. You would be better off ignoring everything they teach you.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#472

Post by cm2187 » Thu, Jun 25, 2026, 6:57 AM UTC

0xbadcafebee wrote: post #11

There's two basic kinds of distillation: 1) the massive [and dumb] method where you ask a question and use the answer as reinforcement (Black Box), and 2) more targeted distillation where you use one model to directly inform/train/guide another model (RLAIF). The latter is basically fine-tuning the model with direction from another model. Thousands of businesses do this every day to fine-tune. This is almost certainl…

Stupid question: I was under the impression that these models were trained on PB of data. Surely the amount of questions/response they can extract from querying a bigger model (Claude) is fairly modest. How is it not a drop vs the training dataset?

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#473

Post by nicce » Thu, Jun 25, 2026, 6:59 AM UTC

tristanj wrote: post #25

Here's what is happening: Chinese resellers are offering Claude tokens at 70-90% below official Anthropic API prices. They achieve this by reselling capacity from pooled Claude Max accounts, payments fraud, and also reselling the model output & reasoning chains to various Chinese labs. They are subsidizing model access in exchange for user logs and reasoning traces, which they then sell as training data, allowing the…

> These resellers operate tens of thousands of bot accounts, which is also why Anthropic introduced identity verification, to slow down the onslaught of bots.

Don’t put that on Chinese.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#474

Post by bozdemir » Thu, Jun 25, 2026, 6:59 AM UTC

Oh wow !!! Antrophic always asks people indiviually if they can train on their personal data. I'm shocked ! Bad Aliba ! Bad....

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#475

Post by gloosx » Thu, Jun 25, 2026, 7:01 AM UTC

So why don't they proceed with a lawsuit instead of public accusations? Let the court decide if these "distillation attacks" are actually illicit.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#476

Post by impossiblefork » Thu, Jun 25, 2026, 7:01 AM UTC

wilg wrote: post #337

tasuki wrote: post #334

> The strike by Alibaba is described as a "distillation" effort, which Anthropic has said involves training a less capable model on the outputs of a stronger one. I don't see what's wrong about this. > Anthropic said the campaign was conducted between April 22 and June 5, 2026, and generated more than 28.8 million exchanges with Claude through almost 25,000 fraudulent accounts. What makes the accounts fraudulent? If…

Because Anthropic has terms of service with more stipulations than just "you must pay and can use the service for any purpose"?

Robots.txt are also ToS of sorts.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#477

Post by zkmon » Thu, Jun 25, 2026, 7:03 AM UTC

I don't understand. If they are simply using our API and paying for tokens, it's called a "transaction" and not "attack". The user is our customer who is supporting our business by buying our services. And we call them attackers. We happily make money by selling our services, and then call it as attack.

Back in the day, an "attack" was supposed to mean be someone acquiring our assets without paying for them or without having our consent. But none of this seems to have happened in this case.

We built a product without paying for most of the raw material we have used, and we don't call that as an "attack". Did we change the meaning of "attack"?

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#478

Post by abbassix » Thu, Jun 25, 2026, 7:04 AM UTC

"It [Anthropic] said DeepSeek's operation involved over 150,000 exchanges". In my humble opinion, a mere 150k exchange for an LLM could only be a benchmarking and not a distillation! I think the US companies should accept that after decades they have rivals surpassing them, just like they did Europeans almost a century ago.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#479

Post by unnouinceput » Thu, Jun 25, 2026, 7:05 AM UTC

Oh, c'mon. If Alibaba wanted, it can have the entire Claude/Mythos source code and data by next week. All you need is enough bribe to a developer that has access to the repository. Humans are always the weakest link in anything.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#480

Post by friendzis » Thu, Jun 25, 2026, 7:06 AM UTC

0xbadcafebee wrote: post #11

There's two basic kinds of distillation: 1) the massive [and dumb] method where you ask a question and use the answer as reinforcement (Black Box), and 2) more targeted distillation where you use one model to directly inform/train/guide another model (RLAIF). The latter is basically fine-tuning the model with direction from another model. Thousands of businesses do this every day to fine-tune. This is almost certainl…

> These complaints of distillation are inflating the problem to make it sound worse than it is

This is, in part, a problem every judicial and legislative system has faced since forever: form versus function.

Take a classic elicitation spying techniques: a foreign spy meets a military officer/scientist at a bar, strikes up a conversation, makes an observation wondering how could a missile hit some target at some accuracy and elicits a response that with laser guidance it is entirely possible. From this they get info that there is some technology to laser guide missiles. Or in retail, a competitor hiring a secret buyer for core baskets of goods and analyzing prices in the receipts.

The function is espionage, the form is conversation and all info is in a sense provided willingly. Where do you pull the slider?

These distillation "attacks" are not only indistinguishable from evals, they ARE evals. The function is own model training, the form is eval. Normally, one would expect to have risk benefit analysis based discussion which direction to push the legality slider to. The problem with these recurring statements is that they invoke enshitification of legislature.