$90–150/hr · micro1
A practicing attorney with corporate, litigation or bankruptcy experience evaluates and shapes AI legal training data by reviewing documents, assessing AI legal reasoning and drafting realistic scenarios.
What you would do
- Analyze and annotate legal documents across corporate law, litigation, bankruptcy and compliance to support AI training
- Evaluate AI-generated legal writing, research and analysis for correctness, precision and adherence to US law
- Draft realistic legal hypotheticals, research memos and case studies that test and develop AI legal reasoning
- Translate abstract legal concepts and procedural rules into structured formats AI systems can learn from
- Identify gaps, errors and emerging risk areas in how AI applies legal doctrine to novel scenarios
Who they want
- JD degree from accredited US law school and active license to practice in at least one US jurisdiction
- 3-15 years of substantive work in litigation, corporate law, bankruptcy and compliance contexts
- Independent practitioner with demonstrated expertise handling complex legal matters
- Strong written and verbal communication skills for explaining legal concepts to non-lawyer technical teams
- Preferred: judicial clerkship, law school teaching, or legal journal editorial experience
Main skills
What the interview asks about
1.Assessing AI legal reasoning accuracy
AI legal writing can sound authoritative while containing subtle errors in doctrine application, case interpretation or procedural understanding; you need to spot these quickly and explain why they matter to practitioners.
For example: “An AI memo on piercing the veil cites three cases correctly but reaches a conclusion inconsistent with law. How would you evaluate the reasoning and what feedback would correct the model?”
2.Drafting training hypotheticals and fact patterns
Realistic legal scenarios must push AI systems to distinguish subtle differences in how doctrine applies across varied fact patterns; generic or oversimplified hypotheticals fail to develop good legal judgment.
For example: “Design a three-part hypothetical about contract formation and express/implied terms, progressively complex. How would you structure fact patterns so AI learns when circumstances matter vs. plain language?”
3.Identifying compliance gaps in AI legal content
AI models often fail to account for jurisdiction-specific requirements, procedural rules and regulatory exceptions that appear in statutory language; spotting these gaps prevents training systems that miss critical compliance issues.
For example: “AI analyzes federal bankruptcy exemptions but ignores state opt-in rules with material creditor impact. What questions would determine whether the model understands when jurisdiction matters?”
4.Annotating complex documents for AI learning
Breaking down dense legal documents requires identifying where key holdings, reasoning and doctrine appear; poor annotation teaches AI systems to miss important arguments or to confuse holdings with dicta.
For example: “A 28-page opinion addresses three issues: two affirmed on summary judgment, one reversed with remand. How would you annotate so AI learns to distinguish holding from reasoning and not apply dicta as binding?”
5.Translating legal doctrine into structured formats
Computational use requires decomposing legal principles into their operative components: conditions, exceptions, consequences, procedural timing; many lawyers think in narrative or analogical ways that don't naturally decompose.
For example: “A statute on piercing the veil lists factors but leaves decisions discretionary. How would you structure this to help AI understand what factors matter, how they interact, and when discretion remains?”
A task you may get
Given an AI legal memo on a corporate law topic (M&A, compensation, compliance), identify three errors or gaps, explain why each matters to a practitioner, and draft corrections.
How to prepare
- Collect three documents (contract, court opinion, compliance brief) from your area. Annotate each, noting key holdings, reasoning, exceptions and judgment calls.
- Draft a hypothetical in your field distinguishing two subtle applications of the same doctrine. Test explaining the difference clearly to a non-lawyer.
- Review one appellate decision in your field and identify where AI models likely misinterpret reasoning, overstate holdings or miss procedural requirements.
- Write a one-page explanation of a complex legal principle (e.g., piercing veil, privity, statutory construction) focused on when rules apply vs. when discretion dominates.
The facts
- Pay
- $90–150/hr
- Open to
- Bangladesh, Hong Kong, India, Indonesia, Japan, Kazakhstan, Kyrgyzstan, Malaysia, Pakistan, Philippines, Singapore, Sri Lanka, Taiwan, Thailand, Uzbekistan, Vietnam, Austria, Belarus, Belgium, Denmark, France, Germany, Greece, Italy, Netherlands, Portugal, Russia, Spain, Switzerland, United Kingdom, Argentina, Brazil, Chile, Colombia, Mexico, Peru, Algeria, Bahrain, Egypt, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Palestine, Qatar, Saudi Arabia, Tunisia, United Arab Emirates, United States, Canada, Nigeria, Kenya, South Africa, Ghana, Ethiopia
- Field
- Law
- Role type
- Specialist
- Posted
- 8/4/2026
- Places left
- 50
We wrote this page from the public micro1 listing. It may be out of date, so read the full posting before you apply.