🚫 Anthropic усложнила копирование Claude в 100 раз и заблокировала десятки тысяч аккаунтов Anthropic заявила, что внедрила новые механизмы защиты Claude от дистилляции — метода, при котором ответы мощной ИИ-модели используют для обучения другой нейросети. По оценке компании, теперь злоумышленникам удается извлекать в 10–100 раз меньше токенов рассуждений Claude, чем раньше. Это затрудняет попытки конкурентов воспроизвести возможности модели без сопоставимых затрат на ее разработку. 🛡Для защиты Anthropic усилила обнаружение подозрительных запросов и ограничила использование прокси-сетей, через которые организовывались массовые обращения к Claude. Компания утверждает, что доля ложных срабатываний системы составляет менее 0,1%. В результате были заблокированы десятки тысяч аккаунтов, связанных, по данным Anthropic, с попытками извлечения данных. Среди предполагаемых нарушителей компания назвала китайские ИИ-стартапы. 📷 BlueScreen | Digital Kazakhstan
Anthropic made copying Claude 100 times more difficult and blocked tens of thousands of accounts
🚫 Anthropic made copying Claude 100 times more difficult and blocked tens of thousands of accounts. Anthropic stated that it has implemented new mechanisms to protect Claude from distillation — a method where responses from a powerful AI model are used to train another neural network. According to the company's estimate,…
AI Summary & Key Takeaways
- 🚫 Anthropic made copying Claude 100 times more difficult and blocked tens of thousands of accounts
- Anthropic stated that it has implemented new mechanisms to protect Claude from distillation — a method where responses from a powerful AI model are used to train another neural network
- According to the company's estimate,…
Anthropic made copying Claude 100 times more difficult and blocked tens of thousands of accounts
Related stories
Recursive has released a model for coding agents
🌟 Recursive has released a model for coding agents The startup Recursive Superintelligence, founded by Richard Socher and emerging from the shadows in May with $650 million in investments, showed its first model - Flint. Recursive received this kind of money even before the first model, largely due to...
Anthropic co-founder Jack Clark moves his prediction for self-improving AI forward to 2027
A new benchmark by P-Zero Research measures how well AI models design and evaluate experiments on their own. The results are improving so fast that Anthropic co-founder Jack Clark now thinks self-improving AI could arrive by 2027. The article Anthropic co-founder Jack Clark…
Sam Altman admitted that no one knows whether artificial intelligence is conscious
Sam Altman admitted that no one knows whether artificial intelligence has consciousness OpenAI CEO Sam Altman said that today no one can say with certainty whether modern neural networks have consciousness. “I want to be clear: in fact, no one is...
News: AI is hacking the Internet. Kimi K3.1 solved 3d, blender and unity
Neural Networks News: AI agents are hacking the Internet: key leaks in OpenCode and Railway, critical vulnerabilities in PHP and GitLab, attacks on banks and a new race between hackers and defenders. Let's look at how AI is changing cybersecurity and why even an isolated environment is not always...