AI Contract Risk Checker

How Clause Risk Scoring Would Actually Work

Not a black box: intake the full clause text, classify each clause by type, tier it High, Medium or Low against common patterns, then return one scored summary instead of a wall of commentary. This page walks the mechanics honestly — including what a risk score can't replace. Not legal advice, and not built as a dedicated tool yet.

Intake · classify · tier · summarize. Not built yet — waitlist below.A first-pass heuristic, not a legal risk assessment
Vote for a risk checker (waitlist)
Clause-by-clause classificationHigh / Medium / Low tieringOne scored summary, not a wall of text

Honest scope

Not built yet — and a risk score isn't a legal opinion

A dedicated structured risk checker (clause classification, repeatable High/Medium/Low tiering, scored summary) is not a VUST bot today. @vustbot can already discuss a pasted clause's risk in chat, but that's a one-off conversation, not a consistent scoring pass across a whole document. The waitlist button is a real demand counter — enough votes and we build the dedicated version. And every score would stay honestly framed: a heuristic flag for where to look, never a certified legal risk assessment.

Everything on this page describes what such a checker would return; the scored pass shown is illustrative, not a live product output.

See the difference

Intake, classify, tier, summarize — the mechanics, honestly, step by step.

Step 1 — intake

What goes in

The full contract text — pasted, forwarded, or (eventually) uploaded as a document. Not a summary of it, not a description; the actual clause language, because risk lives in the specific wording ('may terminate for convenience' reads very differently from 'may terminate for cause').

Why the raw text matters

A risk checker that works from your description of a contract is only as good as your description. Working from the actual text means the same clause gets the same read whether you noticed it or not — the entire point of a checklist over your own skim.

Step 2-3 — classify, then tier (illustrative)

A sample scored pass — not a live product output

Illustrative only, for a vendor services agreement: 'Termination — vendor may terminate for convenience with 15 days' notice, you may not. Tier: HIGH (asymmetric exit rights). Liability cap — capped at fees paid in prior 3 months, standard for this contract size. Tier: LOW. Auto-renewal — 12-month renewal unless cancelled 90 days in advance; short cancellation window. Tier: MEDIUM. IP assignment — all work product assigned to vendor, including pre-existing tools you bring. Tier: HIGH (overbroad scope).'

Why tiering, not just flagging

A flat list of 'here are some clauses' makes you triage yourself. Tiering does the triage: two HIGH items and one MEDIUM tells you exactly where to spend your five minutes of attention, instead of reading the whole document with equal weight.

Step 4 — summarize, and the honest limit

What a summary would return

A short scored list — clause, tier, one-line reason — not a rewritten contract and not a legal opinion. The point is orientation: which two or three things are worth a real conversation before signing, and which parts are unremarkable boilerplate.

What this can't replace

Risk tiers are heuristics based on common patterns, not a certified legal risk assessment. A HIGH tier means 'this is worth a second look,' not 'this will definitely hurt you' — and a LOW tier isn't a guarantee either. For a genuinely high-stakes contract, a lawyer's judgment still beats any tiering system.

02·Practical use cases

Who a risk-tiered pass helps

Anyone facing a long contract

A multi-page services agreement with no obvious place to start reading

A scored triage — which clauses are High risk and worth real attention, which are unremarkable boilerplate.

Small teams without legal counsel on call

Deciding whether a contract needs a lawyer's time at all

A tiered summary that tells you whether this one is routine or has real red flags worth paying for review.

People who've been burned by a boilerplate skim

Wanting a consistent, repeatable check instead of a one-off read

The same clause categories tiered the same way every time — not dependent on how carefully you happened to read that day.

03·How it works

The mechanics — intake, classify, tier, summarize

01Intake the real text

The actual clause language, not a description of it — because risk lives in specific wording.

02Classify each clause

Termination, liability, IP, indemnity, non-compete and more — sorted into known categories.

03Tier and summarize

High/Medium/Low against common patterns, returned as one scored list instead of a wall of commentary.

04·Same tool · in Telegram

Telegram

Vote for a scored risk checker

@vustbot · @vustbot discusses a pasted clause's risk today, free. Press the waitlist button to vote for a structured pass that classifies and tiers every clause in a document.

Open in Telegram

05·Quality & trust

Honest scope — a heuristic, not a legal opinion

Not built yet — this is a demand vote

A dedicated structured risk checker with repeatable tiering is not a VUST bot today. @vustbot discusses a pasted clause's risk in chat now — the waitlist is the vote for the scored, repeatable version.

Not a certified risk assessment

Risk tiers are heuristics based on common patterns, not a lawyer's judgment. For a genuinely high-stakes contract, a lawyer's assessment still matters more.

Frequently asked questions

Related reading.

AI Contract Reviewer

The general output this scoring step would feed into.

Free NDA Analyzer

A narrower, NDA-specific version of the same idea.

AI Lease Agreement Review

The same risk-tiering approach, applied to leases.

Ready when you are

Intake, classify, tier — your vote decides if it ships.

The mechanics are designed; the dedicated tool isn't built. The waitlist is a real demand counter.

Open @vustbotThe general contract reviewer