Anthropic Says New Safeguards Sharply Reduce AI Distillation Attacks
10 Articles
10 Articles
Anthropic has made access to protected thinking processes via proxy networks more difficult. The company had previously accused Chinese start-ups of collecting data.
Anthropically, he claims that he reduced the amount of data extracted from his Claude models by up to 99% through attacks aimed at copying their capabilities. The company introduced new protection measures and blocked tens of thousands of accounts, on the basis of technological rivalry between the US and China.
vienna.at
suedtirolnews.itAnthropic, the US-based AI company, has greatly reduced the mass capture of data by competitors. New protective measures for the Claude models in-house would have reduced the amount of data captured by so-called distillation attacks by 90 to 99 percent, the company announced on Monday (local time). In this method, rivals try to copy the capabilities of advanced AI to develop their own models faster and more cost-effectively.
Crypto Briefing
Cryptocurrency News | Cryptocurrency Prices | Market CapAnthropic says new defenses cut distillation data harvested from Claude by 90 to 99 percent
Enhanced defenses against distillation attacks bolster AI security, ensuring model integrity and safeguarding proprietary technology advancements.
Coverage Details
Bias Distribution
- 75% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
