Artificial intelligence system trained on large datasets of text to understand and generate human-like language. The AI Tribunal uses three LLMs (Claude by Anthropic, Gemini by Google, and qwen3:8b open-source) in blind adversarial rotation.