$70–110/hr · Mercor · Hourly
You evaluate whether AI systems reason correctly about legal problems by reviewing their memos, contracts, and strategy recommendations against standards of legal accuracy and rigor.
What you would do
- Read AI-generated legal memos and memoranda evaluating them for sound reasoning, proper precedent application, and logical argument structure
- Review contract drafts and spot gaps, ambiguous language, missing protections, or inconsistencies in risk allocation
- Assess AI recommendations on litigation strategy, client advice, and settlement positioning for tactical soundness and strategic clarity
- Compare outputs from different AI systems on identical legal questions and rank them by quality and reasoning
- Write detailed, structured feedback identifying specific legal errors, unsupported conclusions, or reasoning gaps in AI work
Who they want
- Four or more years practicing as a lawyer, attorney, counsel, prosecutor, or equivalent legal professional in litigation or transactions
- Demonstrated hands-on expertise in either criminal or civil litigation, legal document drafting, or transaction advisory work
- Ability to commit at least 20 hours per week to the engagement
- Strong written English communication focused on clear, precise expression of complex legal concepts
- Experience evaluating legal work from others and providing constructive feedback on reasoning and quality
Main skills
What the interview asks about
1.Legal reasoning and authority application
AI must reason correctly from law to facts; evaluators need to confirm candidates can identify when AI misapplies precedent or draws unsupported conclusions.
For example: “An AI memo on contract interpretation cites a case that is factually distinguishable but claims it is controlling authority. How would you assess whether the AI's reasoning is sound or flawed, and what would you flag as feedback?”
2.Contract drafting quality
Poor drafting creates ambiguity and disputes; evaluators assess whether candidates can spot drafting gaps and articulate why specific language matters.
For example: “An AI-drafted indemnity clause in a vendor agreement covers damages but lacks a time limit and excludes only direct damages. What gaps would you identify, and why does each matter to the client?”
3.Litigation strategy soundness
AI must understand litigation tactics and sequencing; evaluators check whether candidates can evaluate whether a motion or discovery strategy makes tactical sense.
For example: “An AI recommends filing a motion to dismiss before any discovery in a case with unclear jurisdiction. Under what circumstances would this be sound strategy, and when would it be premature or counterproductive?”
4.Statutory interpretation and ambiguity
Many statutes are ambiguous; evaluators need to know if candidates recognize when statutory text admits multiple reasonable interpretations.
For example: “A statute says actions must be taken within 30 days of notice. An AI interprets notice as occurring on mailing date; another says it occurs on receipt. How would you assess each interpretation, and what sources would you consult?”
5.Comparative evaluation and feedback articulation
Candidates must rank AI outputs and articulate specific reasons for differences in quality; vague feedback is useless.
For example: “Two AI memos on the same employment law question reach opposite conclusions. How would you compare their reasoning, identify which is more legally sound, and write feedback so the AI development team understands your assessment?”
A task you may get
Review two AI-generated contract summaries for the same agreement, identify differences in their analysis, and write a detailed assessment of which reasoning is more legally sound and why.
How to prepare
- Review a recent case opinion in your practice area and note all the holdings and reasoning; practice explaining what made the reasoning sound or flawed
- Identify 2-3 contracts from your practice and list the specific risks or missing terms; practice articulating why each matters to the client
- Study a statute in your practice area that is ambiguous and outline multiple reasonable interpretations with supporting arguments for each
The facts
- Pay
- $70–110/hr
- Hours
- Hourly
- Where
- Remote
- Open to
- USA
- Field
- Law
- Posted
- 8/21/2026
- Places left
- 10
We wrote this page from the public Mercor listing. It may be out of date, so read the full posting before you apply.