MEASURE 2.5

Control

The AI system to be deployed is demonstrated to be valid and reliable. Limitations of the generalizability beyond the conditions under which the technology was developed are documented.

Incidents linked to this control

From automatic embedding match (not a formal assessment).

Racial Bias in Lung Function Diagnostic Algorithm Leads to Underdiagnosis in Black Men

June 1, 2023

Apparent Failure to Accurately Label Primates in Image Recognition Software Due to Alleged Fear of Racial Bias

May 22, 2023

False Arrest of Georgia Man Due to Louisiana Police's Faulty Facial Recognition Technology

November 25, 2022

Stable Diffusion Abused by 4chan Users to Deepfake Celebrity Porn

August 17, 2022

Hiring Algorithms Provided Invalid Positive Results for Interview Responses in German

July 1, 2021

Year-long AI Surveillance Pilot in Two South Australian Aged Care Facilities Reportedly Overwhelmed Staff with False Positives

March 1, 2021

High False Positive Rate by SWP's Facial Recognition Use at Champion's League Final

June 3, 2017

Government Deployed Extreme Surveillance Technologies to Monitor and Target Muslim Minorities in Xinjiang

October 1, 2016

Amazon’s Experimental Hiring Tool Allegedly Displayed Gender Bias in Candidate Rankings

August 10, 2016

Tougher Turing Test Exposes Chatbots’ Stupidity (migrated to Issue)

July 14, 2016

Police Departments Reported ShotSpotter as Unreliable and Wasteful

October 9, 2012

Yes, artificial intelligence can be racist