Ai Training is a topic tracked in our intelligence system with 5 linked articles.
Elon Musk reportedly admitted under oath that xAI has used OpenAI's models to train its own, arguing it's standard practice to use competitors' models.
Meta is deploying a Model Capability Initiative to monitor US employees’ computer activity (mouse movements, keystrokes, occasional screenshots) to train its AI agents, with claims it won’t affect performance reviews but raising privacy and regulatory concerns.
Meta reportedly plans to capture employees' mouse movements and keystrokes to train AI, raising privacy and regulatory risk concerns.
Deccan AI, a Mercor competitor, raised $25M and is concentrating its workforce in India to manage quality in a fast-growing, fragmented AI training market.
Anthropic alleges DeepSeek, MiniMax and Moonshot used Claude to distill training data, creating ~24,000 fraudulent accounts and over 16 million exchanges.
Title asserts 72 million points of interest with only a Hacker News comments link; no concrete details.
Subscribe for real-time topic updates and unlimited access to our intelligence platform.