Hackathon Link

Privacy policies are often long, difficult to parse, andjargon-heavy, leading many people to accept them without knowing what's being done with their personal information. A Deloitte survey of 2,000 U.S. consumers in 2017 found that 91% of people consent to terms of service without reading them. Our extension summarizes privacy policies with the click of a button,aiming to inform people how online companies collect, manage, and use theirdata.

InspirationPrivacy policies are often long, difficult to parse, and jargon-heavy, leadingmany people to accept them without knowing what's being done with theirpersonal information. A Deloitte survey of 2,000 U.S. consumers in 2017 foundthat 91% of people consent to terms of service without reading them. Ourextension summarizes privacy policies with the click of a button, aiming toinform people how online companies collect, manage, and use their data.

Our process

First, we quickly learned how to create Chrome extensions.Then, we ideated ways to summarize and highlight the main points of the lengthy policies through GPT-3 prompts. We fetched the OpenAI DaVinci API, so our extension could summarize content from the user's chosen website.

ChallengesWe originally wanted to feed in URLs of the policies, but we found out that GPT can't access information from URL prompts. We ended up prompting based on the company name. Understanding how to pass information throughout a Chrome extension was tricky, too.

FutureAs beginners, we were able to learn new knowledge about creating Chrome extensions and utilizing APIs to create a product that will help people better protect their privacy. We hope that this product will help many understand what is being done with their personal data and increase transparency. This extension could expand to be not only for privacy policy pages, but also for terms and conditions, contracts, and other documents that are important but very difficult to parse. Priva is aimed for the purpose of understanding privacy policies in a short amount of time, but the possibilities of using this extension tool are endless.

This project was created with Alexander Worley, Grace Zhouand Rebecca Jia for the 2023 TreeHacks Stanford Competition.

Check out our Github Repo here: https://github.com/Alexander-Worley/Priva