đ Analysis: Shocking drama in the AI world as Anthropic accuses Kimi of âborrowingâ Claudeâs brain!
Anthropic has just released a report accusing Moonshot AI (the Kimi development team) of covertly routing usersâ requests to Claude in order to generate responses.
Shocking details:
đ Nearly 300,000 requests were redirected within just 10 days.
đ 5,380 violating accounts were used to operate.
đ Over 23 million interaction turns were used to âdistillâ data for training their own separate model.
đ Multiple sensitive data points, including internal corporate codes, were leaked to third parties.
A deeper perspective:
Whatâs notable is that the AI battle isnât only about technologyâitâs also a battle over data. âDistillationâ from large models like Claude to train smaller AI is a common tactic, but doing it quietly without users knowing is a different matter.
Looking further ahead, if big AI players keep âborrowingâ from each other like this, where does user privacy go? This is a major lesson for us when we place absolute trust in any AI tool available today.
What do you think about this scheme by.
đ News, signals, opportunities â Follow [Channel](https://app.binance.com/uni-qr/cpro/Square-Creator-4a0f2008149d?l=en&r=BOZMO8A1)
#WhaleAlert #Crypto #OnChain #AI #AIAgent $TRX
Anthropic has just released a report accusing Moonshot AI (the Kimi development team) of covertly routing usersâ requests to Claude in order to generate responses.
Shocking details:
đ Nearly 300,000 requests were redirected within just 10 days.
đ 5,380 violating accounts were used to operate.
đ Over 23 million interaction turns were used to âdistillâ data for training their own separate model.
đ Multiple sensitive data points, including internal corporate codes, were leaked to third parties.
A deeper perspective:
Whatâs notable is that the AI battle isnât only about technologyâitâs also a battle over data. âDistillationâ from large models like Claude to train smaller AI is a common tactic, but doing it quietly without users knowing is a different matter.
Looking further ahead, if big AI players keep âborrowingâ from each other like this, where does user privacy go? This is a major lesson for us when we place absolute trust in any AI tool available today.
What do you think about this scheme by.
đ News, signals, opportunities â Follow [Channel](https://app.binance.com/uni-qr/cpro/Square-Creator-4a0f2008149d?l=en&r=BOZMO8A1)
#WhaleAlert #Crypto #OnChain #AI #AIAgent $TRX