Highly Codified Subject Matter Fully Represented in AI Training Data
#1Criminal law and procedure are among the most extensively documented knowledge domains in existence — every statute, court opinion, procedural rule, and criminological theory is text-encoded and heavily represented in LLM training corpora. GPT-4 scored in the 90th percentile on the Uniform Bar Exam in 2023, demonstrating that AI systems have already internalized criminal law at a level exceeding the average practicing attorney, let alone a community college instructor. Models fine-tuned on legal corpora (Harvey AI, CaseText CoCounsel) now provide real-time legal research and doctrinal analysis at a quality level that law firms are deploying in production, creating a public benchmark against which instructor knowledge-transfer value is being implicitly measured.