Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

TL;DR - automate privacy protection in order to serve the uninformed masses of people leaking sensitive data online

We're disclosing more and more private information to assistants, but especially to Facebook, Google and the respective phone company we're using. I think it would be the case to study automatic detection of sensitive information disclosure, in order to place better privacy guards.

I envision a system where the web browser or voice agent would immediately know the sensitiveness level of information we are about to divulge, and route it through anonymous systems or block the leak before it happens.

A database containing our online identities, credit cards, passwords and text run through a topic classifier would make good features for sensitive disclosure protection. It would be the privacy equivalent of antivirus software. Maybe we could have a smart (privacy protecting) web browser and agent.

In the future, people are going to have to convince their personal assistants if they need to disclose any private info to third parties. It's not going to be so easy to collect massive hordes of private data about people. It's just the natural step for privacy, in a world where AI is already being used so much to undermine it. Time to get some AI fighting for our side of the privacy war.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: