Reddit CEO Steve Huffman has publicly criticized tech companies like Microsoft, Anthropic, and Perplexity for scraping Reddit's data without permission to train their AI models. Huffman emphasizes that Reddit's data is not freely available and must be paid for, contrasting with the stance of these companies. Despite Microsoft's claims of respecting the robots.txt standard, Reddit blocked Microsoft and others for non-compliance and unauthorized data use. Reddit has made a licensing deal with Google, allowing it to use Reddit data for AI training for $60 million per year. Huffman expressed frustration over these issues, noting that blocking these companies has been a significant challenge. This move underscores Reddit's broader efforts to control and monetize its data usage by AI firms.