Anthropic logo, updated Anthropic has been aggressively scraping data from websites to train its Claude LLM, permission or no. They don’t care — they want your data. [FT, archive] This isn’t unusual — egregious plagiarism has long been standard in AI model training. But Anthropic have gone pathological. Ifixit.com says Anthropic’s ClaudeBot hit their site a million times in one day: [Twitter, archive] Hey @AnthropicAI: I get you’re hungry for data. Claude is really smart! But do you really need ...