Facebook AI-Supported Moderation for Ads Failed to Detect Violating Content
Dec 8, 2021
Facebook's ad moderation system involving algorithms failed to flag hateful language and violating content such as calls for killings for ad...
Read moreValues Statement: We believe AI should cause no harm, but enhance the quality of human life, by proactively adopting our AI Governance framework.
Evidence-based Transparent For governance
AI Incidents
Data source & citationUK Government AI Allegedly Targets Disproportionate Numbers of Certain Nationals for Fraud Review
Dec 6, 2021
The UK's Department for Work and Pensions (DWP) faced scrutiny after many Bulgarian nationals reported unexplained suspensions of their Univ...
Read moreDeepfake Allegations Raise Concerns Over Trustworthy AI in South Korea
Dec 6, 2021
In an incident that highlights the need for responsible AI governance, a South Korean political candidate used a deepfake avatar, raising al...
Read moreFacebook's Algorithms Failure to Detect Hate Speech Disproportionately Affecting Minority Groups: An Examination
Nov 21, 2021
An investigation by Facebook researchers revealed under-reporting of less common but harmful content towards minority groups, including Blac...
Read moreMisrepresentation of AI-Powered Web Accessibility Solutions: Allegations Against Vendors
Nov 21, 2021
Vendors offering AI-driven web accessibility services allegedly misled customers, claiming their products could deliver automated compliance...
Read moreA Robot at a Tesla Factory in Texas Allegedly Injured an Engineer
Nov 10, 2021
A Tesla manufacturing robot at the Giga Texas factory is reported to have malfunctioned, injuring an engineer. The robot, which was designed...
Read moreDUI Arrest Case Allegedly Based Only on ShotSpotter's Alert
Nov 7, 2021
A car stop resulting in a DUI arrest of its driver was allegedly based solely on a ShotSpotter alert, the reliability of which came into que...
Read moreTesla on FSD Reportedly Drove into the Wrong Lane in California
Nov 3, 2021
A Tesla Model Y in Full Self-Driving (FSD) mode drove into the wrong lane after making a left turn despite its driver allegedly attempting t...
Read moreZillow Discontinues AI-Powered Division Due to Predictive Pricing Tool's Limitations: A Cautionary Tale for Trustworthy AI
Nov 2, 2021
The AI-driven predictive pricing tool Zestimate at Zillow reportedly struggled to forecast housing prices accurately, leading to a three-to-...
Read moreKenyan Data Annotators Allegedly Exposed to Graphic Content for OpenAI's AI
Nov 1, 2021
Sama AI's Kenyan contractors were reportedly asked with excessively low pay to annotate a large volume of disturbing content to improve Open...
Read moreRegulator Suspends Pony.ai's Driverless Testing Permit after Non-Fatal Collision: Ensuring Safe and Secure AI Operations
Oct 28, 2021
The California regulator's decision to suspend Pony.ai's driverless testing permit follows a collision involving one of the startup's autono...
Read moreRacially Biased AI Model: A Case for Responsible AI Governance
Oct 22, 2021
The publicly accessible Delphi AI, allegedly biased in ethical advice due to training via Reddit threads, highlights the importance of trust...
Read moreData source
Incident data is from the AI Incident Database (AIID).
When citing the database as a whole, please use:
McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.
Pre-print on arXiv · Database snapshots & citation guide
We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.