In the United States, thousands of otherwise viable claims routinely
drown in a sea of transaction costs. If you’re a customer wronged by a
company, chances are you have little practical legal recourse. Hire a lawyer?
Good luck—an hour of attorney time might exceed the value of your claim.
Small-claims court? Maybe—but it’s rarely worth the time, effort, stress, and
risk. Class actions aren’t under your control, and anyway, that seven-dollar
check is barely worth the trip to the post office. Then again, none of that is
even on the table: Remember that forced arbitration clause on page thirtynine
of that clickwrap agreement?
Ballooning transaction costs create problems for companies contracting
with one another as well, where the combination of uncertainty, time,
litigation or arbitration fees, and asymmetrical power dynamics routinely
leads to bad outcomes and suboptimal dispute resolution.
But new technologies bring new solutions, and major developments in
artificial intelligence, deep learning, and computational statistics offer a
powerful and elegant way to drastically reduce the transaction costs associated
with resolving a wide array of claims.
This Article draws from advancements and fundamental principles in
the fields of computer science, artificial intelligence, economics, and
medicine to propose and predict a novel, contracts-based private mechanism
enabling parties to voluntarily resolve a wide array of disputes. Parties would
use a flexible AI platform with a known margin of error that they would
accept at the outset, in exchange for a combination of reduced transaction
costs, the chance to resolve a dispute that might otherwise be irremediable,
and finality.
This Article further and relatedly argues that, if successfully
implemented, such a platform could precipitate a rapid shift in how a range
of disputes are resolved and could change the use cases for class actions and
mass arbitrations for the better.
