LO
Get told when a verified record lands on this occupation → · Mark which of these tasks are yours (VOLO Pro, free during the launch) →
Judge / magistrate
Hears cases and runs proceedings, researches the law, drafts judgments and orders, decides outcomes and sentences, and manages a docket. Judiciaries have started putting AI on judges' desks — England and Wales made a private Copilot Chat available to judicial office holders — while writing down where it may not go: judges must read the underlying documents, AI is a poor way to research new information, and courts should not delegate decision-making to it. China's Supreme People's Court says AI must never stand in for a judge's ruling while encouraging it for drafting assistance and case administration. Legal research tools still hallucinated in 17% to 33% of answers in a Stanford study. The US projects judge employment to grow 2 percent to 2035, without mentioning AI.
Read your court's current AI guidance and note which tools it approves and for what.
This is not a probability of losing your job. It combines how much of the role's task load is exposed to automation with how far adoption has actually gone — useful for comparing occupations on one consistent basis, and for nothing else.
Written for judges and magistrates in courts. Lawyers, paralegals and court reporters have their own pages; administrative hearing officers and arbitrators do related but different work. The evidence is judiciaries' own guidance and rules in England and Wales, New Zealand, California, the US federal courts and China, the EU AI Act, a US labour projection and two studies of legal research tools; it establishes what courts permit and forbid AI to do in judges' work, not how many judges use it or how cases are decided in practice.
What is actually changing#
The unit of analysis is the task, not the job title. A role is not replaced — its task mix shifts.
Each tile is one task. Its size is how much of the job it is; its colour is where the task is heading. Click a tile to see what the judgement does not establish.
Guidance from one judiciary and a projection for one country; no record tests AI conducting hearings.
Tests of research tools, not of judges' work, with models that have since changed; guidance says what is advised, not what judges do.
Tools made available and policies supporting them; no record measures how many judgments are drafted with AI.
Rules and guidance, not evidence of what happens in individual cases; the EU obligations for such systems apply only from December 2027.
Policy support and a rule requiring policies; neither shows how much administration is automated.
Is this your job? Say so and this page narrows to your share of it.
A job title is a bundle of tasks bought together, and no two people hold the same bundle. Nothing is sent anywhere — it stays in this browser.
Read all 5 tasks in full — direction, reasoning and limits →
Recent changes#
United States, federal courts. The judiciary's news release on a Judicial Conference meeting says that, under interim guidance from its AI task force, courts have been cautioned not to delegate core judicial functions to AI, including decision-making or case adjudication, and that all judiciary users have been reminded they are accountable for their work. The interim guidance itself is not public; this is the judiciary's description of it.
Failure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
United States. The statistics bureau counts 42,700 judges and hearing officers in 2025 and projects employment to grow 2 percent from 2025 to 2035, slower than average, saying budgetary constraints may limit governments' ability to fill vacant positions or authorise new ones. The page does not mention artificial intelligence, and the category includes administrative hearing officers.
A named person with standing publicly predicted something, on a date, in an attributable statement. It is recorded so that who said what, and when, stays checkable — and it never moves a task's assessment, because a prediction is not an observation. Its value arrives later: the record sits on the same page as the evidence about that occupation, so anyone reading the forecast reads the record of what happened next beside it. That is the reckoning; this site publishes no verdict on whether a forecast came true.
England and Wales. The judiciary's guidance to judicial office holders says they are personally responsible for material produced in their name; that judges must always read the underlying documents and AI cannot stand in for direct judicial engagement with evidence; that AI tools are a poor way of conducting research to find new information you cannot verify; and that nothing not already public should be entered into a public AI chatbot. Guidance, not statute.
Failure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
California. Rule 10.430 says any court that does not prohibit the use of generative AI by court staff or judicial officers must adopt a generative AI use policy by December 15, 2025, applying to the superior courts, the Courts of Appeal and the Supreme Court. It governs how AI is used in courts, not whether judicial tasks move to it.
Failure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
England and Wales. The refreshed guidance introduces a private AI tool, Microsoft's Copilot Chat, which is now available on judicial office holders' devices through eJudiciary, and applies to judicial office holders, their clerks and judicial assistants. It says the tool is available; it gives no figures on use, and the October 2025 version of the guidance no longer has that section.
An employer has put it into production. Can move the baseline — weighted by scale and how similar the setting is.
European Union. Annex III point 8(a) lists as high-risk AI systems intended to be used by or on behalf of a judicial authority to assist it in researching and interpreting facts and the law and in applying the law to a concrete set of facts. Under Article 113 as amended, the high-risk requirements for Annex III systems apply from 2 December 2027. It regulates the systems and their deployers; it does not forbid courts to use them.
Failure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
A study that tested retrieval-augmented legal research tools sold by LexisNexis (Lexis+ AI) and Thomson Reuters (Westlaw AI-Assisted Research and Ask Practical Law AI), and found each hallucinated between 17% and 33% of the time, with substantial differences between systems. It tests research tools used across the legal profession, not judges' work, with 2024 versions.
A demo, benchmark or paper shows the task can be done. Updates what the technology can do — not what employers will do.
A study of general-purpose language models asked specific, verifiable questions about random federal court cases, finding legal hallucinations between 58% of the time with ChatGPT 4 and 88% with Llama 2, and that models often fail to correct a user's incorrect legal assumptions. Older models from 2023, from the same research group as the study of legal research tools on this page.
A demo, benchmark or paper shows the task can be done. Updates what the technology can do — not what employers will do.
New Zealand. The judiciary's guidelines for judges, judicial officers, tribunal members and support staff say using GenAI chatbots for legal analysis is not recommended because generative AI produces text based on probability rather than understanding and cannot critically examine legal questions. Guidance with no force of law, from 2023, before the tools of 2025–26.
Failure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
China. The Supreme People's Court's opinion sets an assisting-trial principle: however far technology develops, AI must not take the place of a judge's ruling, its output may serve only as a reference, and judicial responsibility rests with the judge. It supports AI for evidence guidance, pushing similar cases, assisted generation and review of legal documents, and administrative work such as automatic file classification, case triage between simple and complex matters and automatic generation of hearing records. A policy opinion; it does not measure use.
Failure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
What this means for you#
If you are heading for the bench, expect AI tools in chambers for drafting and administration, and build what every judiciary reserves: reading the evidence yourself, checking every authority, and deciding.
Expect drafting and administration to change first, and to answer for anything produced in your name. The decision stays with you, and the rules say so.
Your options#
Four directions, each with its real constraints and one thing you can test this week. Continuing as you are is a legitimate choice — it just has to be a chosen one.
Use approved tools for drafting and summaries, check every authority
Judiciaries now provide private AI tools while warning that legal research tools hallucinate.
What is approved differs by court, and some courts prohibit generative AI.
Read your court's current AI guidance and note which tools it approves and for what.
Stay in hearing and deciding cases
Every judiciary on this page keeps decision-making with the judge.
Judicial posts are few, and appointment depends on long legal careers and public budgets.
Compare two judiciaries' AI guidance and note where each draws the line on decision-making.
Move into court AI governance
Courts must now write AI use policies, and the EU treats judicial AI as high-risk.
These roles are few and usually go to senior judges or court administrators.
Find out who wrote your court's AI policy and what it requires of judicial officers.
Common questions#
Not on the evidence here. Judiciaries are providing AI for drafting and administration, but every one on this page keeps the decision with the judge: China's Supreme People's Court says AI must never stand in for a judge's ruling, and US federal courts have been cautioned not to delegate decision-making to it. The US projects judge employment to grow 2 percent from 2025 to 2035.
We do not answer that with a number of years. Watch what your judiciary's AI guidance permits and whether it changes what judges must do personally. Those tell you more than any date.
Some courts provide approved tools — England and Wales made a private Copilot Chat available to judicial office holders — while warning that public chatbots can expose confidential information and that AI is a poor way to research law you cannot verify.
Not under the rules on this page. China's Supreme People's Court says AI output is only a reference and rulings are always made by judges, and the EU treats AI that assists judges as high-risk.
What these judgements rest on#
5 of 5 task judgements on this page are backed by a verified event and 0 are platform inference, each labelled where it appears. Behind them sit 2 technology dimensions, a reconstructed trajectory since language models reached the public, and 10 verified events.
See which technologies, how it got here, and the method →
Where it sits in the official classification: skills, knowledge, related jobs →
Other roles in the same function#
A company divides its work into functions before it divides it into jobs. These sit in Legal & compliance alongside this one — a fact about org charts, not a judgement that they are similar or that they are changing in the same direction.