noworkleft.com

Home › Jobs › Administrative Law Judges, Adjudicators, and Hearing Officers

Legal · SOC 23-1021

Administrative Law Judges, Adjudicators, and Hearing Officers

Conduct hearings to recommend or make decisions on claims concerning government programs or other government-related matters. Determine liability, sanctions, or penalties, or recommend the acceptance or rejection of claims or settlements.

Will AI replace this job?

The data cannot say, and anyone who gives you a percentage for “replaced” is guessing. What it does show is a gap between three things, for the 14 tasks O*NET lists for this occupation:

Could an LLM do it?

7%–14%

of tasks: an LLM alone could do the task or cut its time by half. Human raters say 7%, GPT-4 says 14%.

Counting tasks that need extra software around the model: 57%–86%.

Is it happening?

3 of 14 tasks

show measurable use in Claude conversations (Anthropic Economic Index). Anthropic’s occupation score, which counts only tasks that were also rated feasible, is 30%. Observed is a subset of “could” by construction.

Claude only, mostly US users, sampled periods in 2025. Not all AI use, and not employment.

What is left?

1 of 14 tasks

were rated “no LLM speed-up” by every rater (so, by Anthropic’s method, they cannot show up as observed use either). That is the part of the job these studies see no sign of AI touching.

It does not mean those tasks are safe. It means these methods found nothing there.

Task by task

Sorted with the tasks seen in use first. The three coloured squares are the three raters, in this order: human raters, GPT-4, GPT-4 with an alternative rubric. “Usage score” is Anthropic’s published task_penetration value; 0 means none seen.

LLM aloneneeds extra softwareno speed-up
Task (O*NET 31.0)Could an LLM?Claude usage score
Prepare written opinions and decisions. 2 of 3 raters 0.83
Research and analyze laws, regulations, policies, and precedent decisions to prepare for hearings and to determine conclusions. 2 of 3 raters 0.87
Determine existence and amount of liability according to current laws, administrative and judicial precedents, and available evidence. 2 of 3 raters 0.79
Explain to claimants how they can appeal rulings that go against them. all raters none seen
Schedule hearings. all raters none seen
Authorize payment of valid claims and determine method of payment. (supplemental) all raters none seen
Confer with individuals or organizations involved in cases to obtain relevant information. (supplemental) all raters none seen
Conduct studies of appeals procedures in field agencies to ensure adherence to legal requirements and to facilitate determination of cases. (supplemental) all raters none seen
Conduct hearings to review and decide claims regarding issues, such as social program eligibility, environmental protection, or enforcement of health and safety regulations. 2 of 3 raters none seen
Review and evaluate data on documents, such as claim applications, birth or death certificates, or physician or employer records. 2 of 3 raters none seen
Recommend the acceptance or rejection of claims or compromise settlements according to laws, regulations, policies, and precedent decisions. (supplemental) 2 of 3 raters none seen
Rule on exceptions, motions, and admissibility of evidence. 1 of 3 raters none seen
Issue subpoenas and administer oaths in preparation for formal hearings. 1 of 3 raters none seen
Monitor and direct the activities of trials and hearings to ensure that they are conducted fairly and that courts administer justice while safeguarding the legal rights of all involved parties. no rater none seen
How to read this. These are averages over task lists written for the US labour market, rated by people and by GPT-4 using early-2023 model capabilities, next to a count of what one company’s chatbot users actually ask for. A task being “exposed” says nothing about whether employers will cut jobs, shift work, or hire more of you. See how to read an exposure score.

Try it with your own task list

Sources: task list from O*NET® Database (U.S. Department of Labor, CC BY 4.0; modified by noworkleft.com); “could” labels from Eloundou et al., GPTs are GPTs (OpenAI/University of Pennsylvania, MIT licence, labels from early-2023 models); “seen in use” from the Anthropic Economic Index (Claude usage only, CC BY). Dates, versions and limits.

Other jobs in Legal

Data built 2026-11-06. O*NET 31.0. Task match rates: 99.4% with the OpenAI/Penn labels, 99.2% with Anthropic usage.