The Trust Model5 min read

Review Level (LAM): sufficient supervision, made auditable

How much human review is enough for a task? LexCore turns that judgment into an explicit, auditable dial that drives how many findings escalate to you.

"Review Level" is LexCore's implementation of the Lawyer Analysis Metric (LAM) — the question of how much human review counts as *sufficient* supervision of AI for a given task. Supervision runs on a spectrum from review-everything (safe, no time saved) to review-nothing (sanctionable); the right point is task- and risk-dependent.

Review LevelTypical taskEffect
LowRoutine extraction, defined-term pullsHigh auto-pass; few escalations
ModerateStandard contract reviewBalanced; the default
HighUnusual clauses, liability, novel riskLow auto-pass; more findings escalate
HighestDeal-critical, strategyAlmost everything reaches a lawyer

The level sets the confidence bar a finding must clear to auto-pass. The higher the level, the more findings escalate to a Human Gate. Every review records the level applied — turning a hand-waved judgment into a defensible, auditable record.

What LAM is not

Review Level is a supervision policy and record, not a correctness guarantee. A finding that auto-passes is still only as trustworthy as its CitationGate-verified grounding. The dials are policy settings, and they can be calibrated against measured accuracy over time.

Verification Guarantee: LexCore rejects any AI assertion it cannot mathematically trace to character offsets within the source agreement. It serves as a mechanical review assistant and does not constitute formal legal advice. For technical architecture details, see the Security & Trust Center.