WIRED

@WIRED ↗

Deleted No

Hibernated No

Last Checked March 10, 2023

Created

Mon Mar 06 14:32:53 +0000 2023

Likes

Retweets

Source

Twitter Web App

View Raw Data

JSON Data

View on Twitter

Likely Available

WIRED @WIRED

In 2017, the city of Rotterdam deployed a machine learning algorithm to generate a risk score for around 30,000 people receiving welfare payments. The algorithm used attributes such as age, gender, and Dutch language ability to determine who should be investigated for fraud. 2/10 — PolitiTweet.org

Posted March 6, 2023

Preceded By

WIRED @WIRED

What if your government gave an algorithm the power to assign you a “risk score”? What if the risk score could determine the course of your life? What if that algorithm was full of flaws and biases? Welcome to the suspicion machine https://t.co/C8YQoQLlH6 📷 @katherinelamart https://t.co/IrHvskoEH8 — PolitiTweet.org

Posted March 6, 2023

Followed By

WIRED @WIRED

WIRED and @LHreports obtained Rotterdam’s welfare fraud algorithm and the data used to train it. By reconstructing the system, we found that it discriminates based on ethnicity and gender and has fundamental flaws that make it both inaccurate and unfair. 3/10 — PolitiTweet.org

Posted March 6, 2023