The World’s First Robot Lawyer
DoNotPay leverages AI to help the average consumer compete legally with Governments and Major Corporations
Origin Story
DoNotPay was originally created to solve a personal problem experienced by its founder. While attending Stanford, Joshua Browder repeatedly earned parking tickets for forgetting to feed his meter. A friend suggested that he contest the tickets to have them reduced. He wrote a letter to the city and his fine was reduced. This seemingly minor interaction exposed Browder to a larger truth: there are dozens of fines, tickets, contracts, and terms that the average consumer could contest, but doesn’t because they either don’t have time or because it’s not financially feasible to use a lawyer to do so. Browder set out to fix that problem, and DoNotPay was born.
What is DoNotPay?
DoNotPay brands itself as “the world’s first robot lawyer” and promises to help customers “fight corporations, beat bureaucracy, and sue anyone at the press of a button”
DoNotPay has built a series of discrete integrations with airlines, websites, governments, and lawyers all with the intent of making it easier for the average consumer to pursue legal claims that would’ve previously been financially infeasible to pursue.
For example, let’s say you receive a parking ticket while visiting your parents in your hometown for the holidays. You may disagree with the ticket, and have a good chance of getting the fee reduced if you were to show up, but you don’t have time to represent yourself in court and the return on investment (potential ticket reduction X probability of success / cost of representation) doesn’t make sense. Previously in this scenario you would’ve paid the full ticket price and moved on. Now with DoNotPay you can fight that ticket for just $36 per year. Already DoNotPay has helped customers save an aggregate of more than $20m fighting excessive parking tickets.
How does it work?
A consumer who wishes to contest a ticket, request a refund, cancel a hard to get out of subscription, or fight persistent email or phone call spam can sign up for DoNotPay on an annual subscription of $36 per year.
Once they’ve signed up they can use DoNotPay’s AI powered lawyer technology to contest their claim. The user journey is fairly straightforward:
- The user inputs their basic information.
- The user inputs the details surrounding the specific issue they’re looking to contest.
- DoNotPay ingests that information and uses Artificial Intelligence to transform the inputs into a legal document. Depending on the issue being contested this could be as simple as a strongly worded email, or as formal as a notarized court petition.
- DoNotPay uses AI to respond to any counter arguments or objections from the party being sued or challenged.
- DoNotPay passes along the savings they’ve recovered back to the customer in their entirety.
DoNotPay provides this kind of service across a variety of different claim types. The most popular ones are:
- Tickets & non-moving violations
- Service & subscription cancellation
- Refunds for poor service (Airlines, Rental Cars, Hotels, etc.)
- Spam (email harassment, unwanted robocalls)
How does the AI create value?
The DoNotPay robot lawyer service creates value in a few key ways:
- It helps people access refunds, reductions, and cancellations that they are entitled to but do not currently have the resources or know how to access.
- It reduces the cost of adjudicating small claims which helps both claimants and companies reduce the amount of money they spend on legal fees.
- Over time the DoNotPay robot lawyer uses AI to get better at arguing cases and pursuing legal recourse, helping customers to win more cases than they would’ve even with a credentialed human lawyer.
How does DoNotPay capture value?
The biggest critique of DoNotPay is that they do not do enough to capture the value they are unlocking with their highly efficient and effective AI. DoNotPay charges a flat subscription fee, instead a per interaction success fee, because they want to come across as an independent and uninterested party in contrast to the performance fee heavy legal market they are replacing. In my opinion DoNotPay would be able to capture more value if they shifted to the industry standard bounty model where they received a fixed percentage of all dollars recovered. It will be interesting to see how their fee structure evolves over time.
Looking Ahead
DoNotPay is in the early days of using AI to help level the legal playing field between consumers and large businesses or governments. There will be two key trends to watch as the business matures:
- New features: Already DoNotPay has built on their original product offering with a feature called “DoNotSign”. DoNotSign helps customers get in front of sticky legal situations by using AI to read terms and conditions before consumers sign up for a product or service. It identifies onerous terms and populates a letter that consumers can send to businesses asking for tweaks. This helps provide customers with more security up front, and avoids the need for small claims litigation down the line. This is just one example of the incremental value DoNotPay can provide to its users as its AI model improves.
- Business Response: While DoNotPay likely frustrates businesses it is not operating on a scale where it’s worthy of meaningful consternation. As DoNotPay continues to grow it may have to contend with businesses re-writing their terms, conditions, and policies to make it more difficult for DoNotPay to sue. This could pose a threat to their overall business model.
Sources:
https://www.theverge.com/2019/11/20/20973830/robot-lawyer-donotpay-ai-startup-license-agreements-sign-arbitration-clauses
https://autom.io/blog/5-lawyer-bots-you-can-try-now
https://www.protocol.com/donotpay-app
https://donotpay.com/
Such a great post Merrill. I can think of numerous times I would have wanted to use DoNotPay in a refund or cancellation case. But I 100% agree that the subscription model would deter a lot of people from making use of the service.
How much of the product operates as fully automated AI vs needing human oversight? What were some of the challenges in training the AI to recognize the various types of claims in order to create the argument for the user?
Definitely a field that needs more automation as the services are so cost prohibitive for people that really need them.