A little known chatbot that has been helping US citizens fight parking tickets has found a new purpose. Donotpay, a chatbot on Facebook Messenger, has been redesigned to help asylum seekers with navigating applications to Canada, the UK, and the US.
Designed by a Stanford university student, Joshua Browder, Donotpay was originally intended to help people get out of parking fines. Which is apparently a problem in the US where it was first deployed. It uses a simply conversational style messaging to identify the problem and help the user fill in the necessary forms. The chatbot has successfully contest some 160,000 tickets.
Expanding Donotpay to asylum seekers took a little extra effort than trying to contest parking tickets. Browder had to contact and work with lawyers from all three countries to figure out how the application process works in each of the countries. He also spoke with successful refugees to figure out how to approach the problem.
Donotpay uses simple English, and tries to distill the often opaque language contained in the asylum documents. It tries to determine the kind of asylum that the applicant needs and directs them to the appropriate form to fill out. The chatbot then also helps the applicant to fill out the necessary details.
Once the process is over, Donotpay deletes all saved data after 10 minutes.
There’s no telling how effective Donotpay will be at helping refugees, and it could be several years before any meaningful data is collected. Most countries have long processing times that could run into several years before deciding on an individual case.
That being said, Browder is already working on translating Donotpay into other languages. Work on an Arabic version has already started.
Browder had originally hoped to launch this version of Donotpay six months ago, but ran into delays while tweaking the system for maximum efficiency. He also decided to add Canada as a possible location at the last minute; due to the political climate in the US at the moment.
[Source: The Guardian, Donotpay]
