attorneyrl.com / coming soon / environment v0

Reinforcement
learning for
legal AI.

AttorneyRL is a reinforcement learning environment where AI agents work simulated legal matters — dockets, deadlines, discovery, motions — and earn reward for outcomes a lawyer would defend, not text that only sounds right.

episode
0
step
0
explore ε
—
return
0.00
last outcome
—
Move your cursor — you add reward. Watch the agent chase it instead of the goal. That is reward hacking, the problem this environment exists to train out.
§ 01 — the environment

Litigation, posed as a Markov decision process.

Every matter is an episode. The agent reads the record, picks a move, and the environment answers with the next state of the case — and a reward computed from rules, not vibes.

s

Observation

The state of the matter as a lawyer would see it on a given day.

  • pleadings & docket entries
  • governing rules & local rules
  • deadlines & the calendar
  • produced documents & privilege log
a

Action

Structured, auditable moves — each one checked before the state advances.

  • file · serve · respond
  • move · oppose · object
  • request · produce · withhold
  • negotiate · settle · appeal
r

Reward

Scored by verifiable checks on the outcome, with hard penalties for the mistakes that end careers.

  • favorable, defensible outcome
  • deadlines met, rules satisfied
  • missed deadline · waived right
  • fabricated citation · sanctions
§ 02 — watch it learn

Reward goes up. Sanctions go away.

Two live simulations, running in your browser right now. Left: episode returns from the agent in the background as its exploration decays. Right: a toy matter graph, walked one move at a time.

Episode return · moving avg (20)simulated · live
Matter graph · one trajectoryG = 0.00
§ 03 — the interface

Familiar API. Unfamiliar stakes.

A Gym-style loop your training stack already speaks. Swap the jurisdiction, the matter type, or the reward model — keep the loop.

import attorneyrl env = attorneyrl.make("FedCivPro-v0") obs, info = env.reset(seed=7) while True: action = policy(obs) # e.g. File(MotionToDismiss, ...) obs, reward, done, trunc, info = env.step(action) if done or trunc: break info["audit"] # every check, every rule cited

Preview of the planned interface. Names may change before release.

§ 04 — the business

Two sides. One exchange for legal intelligence.

The labs building legal AI need data that only lawyers can produce. The attorneys who hold that expertise have no way to sell it. AttorneyRL connects them.

For AI labsBuy

Buy legal data that holds up.

Training and evaluation data built with practicing attorneys — scoped to the jurisdictions and practice areas your model needs.

  • 01
    Licensed legal datasetsBy jurisdiction, court and practice area
  • 02
    RL environments & graded trajectoriesEpisodes scored against the rules, not vibes
  • 03
    Preference & reward-model dataPairwise judgments from licensed attorneys
  • 04
    Expert evalsBenchmark your model on real legal judgment
Talk to us about data
For attorneys & firmsEarn

Get paid for your legal judgment.

Licensed attorneys author tasks, grade model outputs and write the rubrics that train legal AI. Paid per task, with a guaranteed hourly floor.

  • 01
    Expert annotationStructure and label legal work in your practice area
  • 02
    Expert evaluationsGrade model outputs on the work you know best
  • 03
    Rubric authoringDefine what a defensible answer looks like
  • 04
    Firm data programComing later, counsel-reviewed, with client consent
Join the attorney network
§ 05 — early access

Train agents that can stand up in court.