Lawyer

$80–180/hrrate as listed

micro1 recruits practicing and formerly practicing lawyers to evaluate AI-generated legal reasoning, drafting, and research output. Work is contract-based, remote, and paid hourly at an observed $80–180/hr depending on jurisdiction, practice area, and screening outcome.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

The work

You review model outputs on legal tasks and judge them the way you would judge a junior associate's memo. In practice that means comparing two responses to the same prompt and deciding which is stronger, rewriting a flawed answer into one you would be willing to sign, or writing prompts and reference answers in your practice area from scratch. Tasks lean toward the areas where models fail quietly: mischaracterized holdings, hallucinated citations, statutes quoted in a superseded form, jurisdictional confusion, and advice that is fluent but wrong on the element that actually decides the matter.

Most items come with a rubric and a required rationale. The rationale matters as much as the score — annotations that say "incorrect standard of review, should be de novo per [cite]" are useful; annotations that say "answer is bad" are not. Some projects ask for adversarial prompts designed to surface a specific failure mode, or for structured redlines of contract language.

What the platform screens for

A verifiable license or clear record of practice. Bar admission and jurisdiction, or a documented history in-house, at a firm, in government, or in academia.
Depth in a named practice area rather than general familiarity. Screens follow up on specifics: elements, tests, controlling authority, common drafting traps.
Citation discipline. Whether you can identify a fabricated or misapplied citation and explain precisely what is wrong with it.
Written reasoning under time pressure. micro1's screening is AI-led and conversational; expect follow-up questions that push on the answer you just gave.
Calibration. Willingness to mark a polished answer as wrong, and to distinguish a stylistic preference from a substantive error.

Logistics

Fully remote and asynchronous, with tasks pulled from a queue. Contributors commonly commit 5–20 hours per week; some projects request a minimum weekly volume or a short window of overlap with a US time zone for calibration calls. Engagements are project-scoped and volume fluctuates — treat this as supplemental contract work, not a replacement for a practice. Pay is quoted per hour and varies by project difficulty and specialization; the band above reflects observed rates, not a guarantee.

What they look for

Drafting and reviewing legal documents
Basic legal research
Identifying risks/problematic clauses (contract review)
Handling confidentiality and conflicts of interest
Explaining legal concepts to non-lawyers

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Are you currently licensed to practice law in at least one U.S. state or territory, or do you hold an equivalent credential in a common-law jurisdiction and have a documented record of legal practice?
Bar admission · required
Can you name a specific practice area (e.g., contract law, employment law, securities regulation, IP licensing, family law) in which you have at least two years of direct experience—research, drafting, advising, or litigation?
Depth in a named practice area · required
Have you worked in a context where you were responsible for verifying legal citations—either as a research attorney, in-house counsel, firm associate, or law clerk—rather than only writing from your own knowledge?
Citation discipline · preferred
Do you have an active, unencumbered license with no history of bar discipline, complaints, or sanctions in the past five years?
Clean record · required
Have you produced detailed written legal analysis (memo, brief, contract review, or annotation) on a short timeline—48 hours or less—with accuracy a priority?
Written output under deadline · preferred

0 of 5 answered

Included with this listing

Interview prep — 11 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether you have the habit and skill to verify authority rather than trust fluent prose, and whether you can articulate the precise error instead of just sensing something is off.
A strong answer: A concrete example where you traced a citation to its source, found it said something different or didn't support the proposition, and explained the specific mismatch—e.g., 'the memo cited Smith v. Jones for the burden of proof, but Smith actually addresses standing, and the actual holding is narrower than the argument required.'
Where people lose it: Giving a vague example ('I found a cite that was wrong') without walking through what the document claimed, what the authority actually said, or why they differed. Answers that amount to 'I have good attention to detail' signal you may not actually verify cites in practice.