AI companies of every kind are looking for data for information and to guide their AI “decision” making process.
This includes legal research and AI solutions envisioned to be used by lawyers, legal academics and the judiciary.
OpenAI, which runs ChatGPT, scrapes the Internet to get a corpus, a metaphysical body, or a collection, of data to train its AI solution. The corpus is the material the AI reviews to become “intelligent.”
Legal research and AI companies scrape the Internet, the law and legal publications to get their corpus. Code, constitutions, regulations, treatises, law reviews, restatements of the law, legal blogs and more.
There’s a question as to what is cited today. Legal citations have always been a prerequisite in the law.
Using ChatGPT, I see citations, and their accompanying links, to a some of the information it shares, but they are few and far between.
Not doing legal research on legal AI platforms, I am not sure what type of citations they use when you may run a document, brief or question through it for feedback. Assuming this is one way which legal AI solutions work.
Got me to thinking whether data will become more important than the law at some point – at two levels.
One, the data is needed to accelerate the way we practice law and adjudicate today. AI gets legal professionals more information and insight, and at lightening speed, as compared to personal research by browsing and search.
Second, is it possible AI will tell us at some point the answer, itself, by virtue of its corpus of the law? Tell lawyers what to write and say? Tell judges what to decide?
All from “data” rather than the law and reasoning, as we know it.