Early bird prices are coming to an end soon... ⏰ Grab your tickets before January 17

This article was published on November 24, 2020

DoNotPay, the AI lawyer for your inbox, now lets you report businesses for tax fraud

Plus, browser integration is on the way


DoNotPay, the AI lawyer for your inbox, now lets you report businesses for tax fraud

Here’s a cool one: DoNotPay, the legalese-trained email crawler that scours your inbox for ways to save you money (or earn you payouts from class action lawsuits), now helps you collect government tax fraud bounties. Natural language processing just keeps getting better.

The software’s pretty simple to use. You open the DoNotPay client, input the details of the fraudulent business including any evidence you can provide, and hit send. The IRS (for US users) or HMRC (for UK users) will then contact you concerning any cash rewards. In the case of the IRS, reporters could be eligible for up to 30% of the total value of assets recovered by the US government.

This is all accomplished through some amazing feats in the field of natural language processing. The same basic principles and technology governing the company’s email offerings power its new “report tax fraud” feature. In describing these AI systems Joshua Browder, DoNotPay’s CEO, told Neural:

The technology in our email system is broken down into two parts: (1) An Email Matching System, and (2) An Auto Responder.
First, Our email matching system relies on a multi-class classification model, where the final prediction will return to the user a set of actions he can take depending on the email that was received.
For example, if you receive an email from United Airlines regarding a Wi-Fi purchase, the top action may be to pursue an in-flight Wi-Fi refund. At our software’s core, we’ve aggregated our training data with thousands of labeled emails and the respective actions users typically take (bank fee appeals, cease and desist notices, subscription cancellation, etc), and after cleaning our data set, we’ve been able to predict with 99% accuracy the most appropriate action a user can take. Aside from analyzing the body of the email, our classification model takes into account other features such as sender, subject line, hyperlinks, etc.
Aside from labeling emails in your inbox, our email service also includes an auto responder. Using this pre-trained model, our system automatically parses intent from incoming emails, and drafts an auto reply that the user can choose to send, or edit freely.
Overall, we’ve been able to leverage cutting-edge developments in Natural Language Processing and Understanding, and have been hard at work incorporating these features for usability in streamlining consumer service correspondence and legal accessibility.

Aside from just operating as a portal for tax fraud whistleblowers, DoNotPay’s main bread-and-butter is its email “robo lawyer.” Both are included in the company’s subscription service.

For $3 a month, in addition to the tax fraud reporting service, you can get an @DoNotPay email address that you can either use to sign up for services in the future or forward all your related emails to. DoNotPay’s AI system will scan your emails for opportunities to save or make you money by cancelling subscriptions, joining class action spam lawsuits, checking your receipts for tax write-offs, and myriad other services including flight check-in and automated refund requests.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!

Browder also told Neural that DoNotPay was currently working on browers plug-ins for Chrome and Safari that will allow seamless integration between their system and your email — something that should allow the AI to crawl your old emails automatically.

Can you imagine what kind of money is lying around in decades worth of emails entombed in billions of forgotten spam folders?

For more info check out the company’s website here.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


Published
Back to top