Computational law

Content sourced from Wikipedia, licensed under CC BY-SA 3.0.

Computational Law is the part of legal informatics that aims to automate legal reasoning. Its key feature is autonomy: these systems can answer legal questions with little or no human input. Today, the main use is compliance—computer systems that check, apply, or enforce rules. You can see this in tax software and other compliance tools. The field is growing fast because of more connected technology and the spread of smart devices.

Computational Law isn’t limited to government rules. It can cover the terms of contracts, insurance and real estate deals, corporate policies, and even the rules of games. Some people imagine machines that can do parts of a lawyer’s job.

The idea goes back to the mid-20th century. In 1949 a field called jurimetrics formed to apply scientific methods to law, including counting decisions and predicting outcomes. In the late 1950s, Lucien Mehl proposed two kinds of machines: a data machine that retrieves legal information and a consultation machine that could answer questions across many areas of law. By the 1970s the information-retrieval machine was common, but fully automating law remained difficult.

In the 1970s and 80s researchers built expert systems that could infer legal conclusions, ask users for needed details, and show the supporting rules. These ideas helped push the field forward. The first AI-and-law conferences began in the 1980s, and the 1990s and 2000s saw rapid growth. Universities started teaching the topics, and a core group at Stanford advanced the work. By the 2000s, more lawyers were studying how automation could change the profession.

Today many see clear benefits: easier self-help with contracts and planning, better prediction of how rules might change, and new automated decision-making tools. Some governments are even testing automated dispute decisions, with human review available on appeal.

There are efforts to make laws machine-readable (structured data) and machine-executable (where a system can run a case to a decision). Machine-readable formats like METAlex are used in the UK and the Netherlands; a 2013 U.S. executive order pushed public documents to be machine-readable.

Current research often analyzes how laws and decisions relate using large networks of citations. Visualizations help people see big patterns, but making these networks readable is challenging because they can be very dense.