Read every clause, flag the risky ones, and show your work.
ClauseProof takes a contract you have to sign (a vendor MSA, an employment agreement, an NDA, a licence) and reviews it clause by clause, telling you where the risk is for your side, why each risky clause actually matters under your governing law, and how you might redline it. Not one unexplained verdict, a per-clause report with the reason attached to every finding.
Book a ClauseProof review See a sample redlined contract
THE PROBLEM
You sign more contracts than you can send to a law firm, and most of them look fine. But the risk never lives in the whole document, it lives in a few specific clauses (the uncapped indemnity, the liability cap that is a tenth of the deal value, the auto-renewal with a ninety-day notice window, the IP assignment that quietly reaches your background IP), and those are exactly the lines that a quick read glides past. The fashionable fix, dropping the whole contract into a chatbot and asking "is this risky", gives you one confident paragraph with no way to check it, which is the kind of answer that feels helpful and falls apart the moment it matters. As of August 2, 2026, the EU AI Act turns "prove what your AI did" into a board-level obligation for high-risk systems, and AI used to review contracts sits right in that scrutiny, so an unexplained AI verdict stops being something you can lean on and starts being something you have to defend.
HOW IT WORKS
ClauseProof does not replace your lawyer and it does not ask you to change how you work, it reads the contract you already have. It segments the agreement into individual clauses, classifies each one into a clause taxonomy (indemnification, limitation of liability, termination, IP assignment, confidentiality, non-compete, payment, governing law, and the rest), and then scores each clause against a risk rubric tuned for your side of the table (buyer or seller, employer or employee, licensor or licensee) and for your jurisdiction (a non-compete reads very differently in California than in New York or the UK). For every clause it flags, it drafts recommended redline language, and it rolls the whole thing into an executive risk summary with an audit trail underneath. It fails closed on purpose, an unparseable contract is never handed back as a clean bill of health, and a clause it genuinely could not assess is reported as unassessed rather than quietly marked low-risk.
WHAT YOU GET (the risk report)
For every clause that matters, one line you can actually check: - the clause and its category, pulled straight from the contract, - a risk level (low, medium, high, critical) for your side, with the reason it is risky, - a suggested redline, drafted for your perspective and your jurisdiction, - an executive summary that reads top to bottom, so you see the shape of the deal at a glance, - an audit trail, so every finding traces back to the input contract and the policy that judged it.
A review a human reviewer can check clause by clause, not a verdict you have to take on faith.
WHO IT'S FOR
In-house counsel, ops leaders, and founders who sign more contracts than they can send to a firm, and legal-ops or contract-lifecycle platforms that want an explainable review layer under their intake. Fractional-GC and boutique legal shops that triage inbound contracts at volume fit too. If a risky clause you missed can cost you real money or lock you into a bad deal, this is for you. If you are in a bet-the-company negotiation where only a named attorney's judgement will do, ClauseProof is your fast, defensible first pass, not the substitute for that attorney.
PRICING (the ladder)
We price the ladder, not a single number, and every number below is a hypothesis we validate with you, not a commitment. - Pilot, a fixed-price paid proof on your own contracts. You see the per-clause report and the redlines on real agreements before you commit to anything. - Per contract, once the pilot proves the value. Your intake keeps running, we review each contract that comes in. - Platform, license the clause-review layer into your CLM, legal-ops, or intake stack.
IP is licensed, never assigned. The reviewer stays ours, the report is yours. Any step that takes your money is gated and confirmed before it runs, nothing charges silently.
THE PROOF (dogfood)
We run ClauseProof on ourselves before we sell it. The same clause-by-clause review is pointed at AYA's own inbound agreements (the MNDAs, the pilot agreements, the vendor MSAs we sign), so risky terms get flagged for the founder before signature, and every run leaves an audit trail, the same shape of proof this product sells. Under the hood ClauseProof is not a special new engine, it is a composition of AYA's real, executable building blocks (extract, classify, analyze, generate, summarize, and an aggregated log for the trail), running through the exact composite path our live demo processes already run through, so eating our own dog food here costs us no new machinery.
HONEST NOTE
Two things are true and worth saying plainly. First, the quality of a review is only as good as the clause taxonomy and the risk rubric we feed it, and those are first-draft policy DATA that we intend to tune with a pilot partner on real contracts, not settled law. Second, this bundle is structurally built and checked offline (the deliverable composes six real on-disk atoms, every reference resolves, and it passes our pattern-quality scorer at grade B), but it has not yet been run on a live model against a real contract, that is the natural first live test. Nothing here has been billed, published, or run in production, the pricing figures are hypotheses, and ClauseProof is an explainable first-pass tool, not legal advice and not a replacement for a licensed attorney on high-stakes matters.
*This page is a specification. The capability it describes is not built yet, and nothing here is a claim that it runs today.*