A legal team at a multi-billion-dollar multinational handles a large volume of cross-border contracts, with long manual review cycles, risk findings that could not be traced, and standards that varied by reviewer. Amway built a unified legal knowledge brain on MOI that flags risk clauses, missing clauses and inconsistent definitions in seconds.
Amway is a leading global direct-selling company operating in over a hundred countries and territories, with annual revenue in the billions of dollars. Its legal team handles a high volume of contracts drawn from many jurisdictions.
Amway's enterprise AI foundation spans contract review, sales enablement, data Q&A and deep research. Contract review went first, because its pain was the most concrete and its return the easiest to measure.
A cross-border commercial contract routinely runs to a hundred pages or more. Legal has to compare each clause against the standard template, check whether required clauses are missing, and verify that a term defined in one section means the same thing in another. The time this takes is set by the contract's length, not its difficulty.
Review cycles ran to 16 days, and from the business side those 16 days were unpredictable: once a contract went to legal, there was no visibility into when it would come back or which clause it was stuck on.
Traceability was the second problem. When legal concluded a clause carried risk, the basis was often individual experience rather than a rule anyone could look up — so a different reviewer could reach a different conclusion.
The third problem was the deepest: experience did not accumulate. A senior lawyer's instinct for risk is among the most valuable assets a company holds, and it was neither quantified nor transferable. When someone left, the judgement left too.
Amway deployed an AI contract review agent, building a unified legal knowledge brain on MOI: standard templates, the required-clause library, historical contracts and prior review opinions are parsed into the store, with clauses embedded for semantic search.
An incoming contract is parsed across formats, its standard clauses and risk points extracted, and then compared three ways in parallel — against the standard template to catch unusual edits, against the knowledge base to catch missing required clauses, and against historical contracts to catch inconsistent definitions.
That triple comparison is the core of the design. Reading one contract in isolation cannot tell you what is absent from it; only placing it back into the frame of template, knowledge base and precedent makes omissions and deviations visible.
Each risk found carries a detailed explanation and a suggested compliant clause, in Chinese or English. Extracted key fields flow automatically into the CLM/ERP system with renewal reminders set, so the review lands in the business process rather than stopping at an opinion memo.
Review cycles compressed from 16 days to the order of a minute, and the business can finally predict when a contract will clear.
AI risk identification reaches 99.9% coverage, and every finding traces back to the template clause or precedent it was compared against, rather than resting on an unverifiable judgement call.
Legal knowledge is now fully captured: senior lawyers' judgement is fixed as rules and examples in the knowledge base, newer staff work to the same standard, and turnover no longer takes the judgement with it.
The role shifts accordingly — from reading and comparing page by page, to ruling on the clauses the system has flagged as genuinely needing a decision.