Anthropic revealed that Chinese AI companies Alibaba, Moonshot, and DeepSeek engaged in unauthorized "distillation" of its Claude AI models, using outputs from Claude to train their own AI systems without permission. The largest operation involved Alibaba, which conducted over 151 million exchanges with Claude between May and July 2026, aiming to enhance its Qwen models. Moonshot secretly routed user requests to Claude, misleading customers about the AI they were interacting with, while DeepSeek employed similar tactics. These activities raise privacy and legal concerns, as some exchanges contained sensitive information. Anthropic disrupted these efforts and reported the findings to highlight misuse risks in AI development.
In the second quarter, the Harbor International Compounders ETF posted a 6.40% return. The biggest drag on performance came from the Information Technology sector, which surged due to a semiconductor rally, causing relative underperformance. Key detr...
Alibaba is moving aggressively from traditional e-commerce to a comprehensive AI and cloud computing strategy, supported by significant insider buying. Its AI Cloud and Compute Services revenue grew 45% year-over-year, with adjusted EBITA soaring 133...
