AI Incident Tracker

Timeline: Subdomains

Insights

Within risk domain '1 Discrimination and Toxicity', the proportion of incidents reported that have been classed as '1.2 Exposure to Toxic Content' has increased since 2020 whereas the proportion classed as '1.1 Unfair Discrimination and Misrepresentation' has decreased.
The number of incidents in domain '6 Socioeconomic & Environmental' is very low relative to other domains, with no more than 3 reported in any year.
The overwhelming majority of reported incidents within domain 7 'AI System Safety, Failures & Limitations' are classed as '7.3 Lack of Capability or Robustness'
Most incidents attributed to Malicious Actors in the last 3 years are frauds, scams or targeted attacks, with very incidents classified as cyberattacks.

‍

Interactive

Detailed view of how AI incidents have changed over time (2015-2025), across the 24 subdomains of the MIT AI Risk Repository Domain Taxonomy

Explore more of the AI Incident Tracker Project

You can explore different views of the database and classification in the project. For example, you can see all AI incidents classified using taxonomies from the MIT Risk Repository, the type of harm, and individual records in the AI Incident Database.

Key visualizations include bar charts and pie charts that display incident counts, proportions across domains (e.g., "System Failures," "Discrimination & Toxicity"), and trends in causal attributes. Additionally, insights highlight patterns such as the prevalence of system safety issues, intentional misuse trends, and incomplete reporting gaps.

Click through the links below to explore each of the interactive dashboards.

Risk Classification Overview

This interactive shows how incidents in the AI Incident Database are classified using the MIT AI Risk Repository’s causal and domain taxonomies, along with risk levels defined by the EU AI Act.

Incident View

View detailed analysis of each AI incident in the incident database, including how the incident is classified, how severe the harm is (across 10 categories of harm), and the classification confidence.

Timeline: Risk Classification

View how AI incidents have changed over time (2015-2024), using Causal and Domain Taxonomies from the MIT AI Risk Repository

Timeline: Subdomains

Detailed view of how AI incidents have changed over time (2015-2025), across the 24 subdomains of the MIT AI Risk Repository Domain Taxonomy

High Severity Incidents

View how AI incidents have changed in severity over time

High Severity Multiple Categories

View the highest severity AI incidents that have resulted in multiple types of harm (e.g. physical harm AND property damage)

Number of People Affected

View the number of medium and high severity incidents where the number of people either harmed or exposed to risk exceeds 1000 and 1 million people.

Harm Taxonomy

View a harm taxonomy for AI incidents, which distinguishes between 10 types of harm (e.g., physical harm, property damage, financial loss, human rights) based on work by CSET

Alleged Developers

View which developers are associated with the most incidents, with different harm severities, and see how the distribution has changed over the past 10 years.

Harm Severity

View how the numbers of incidents causing different levels of harm in each category has changed over time

NatSec Impact Framework

View incident impact over 5 categories and with threat characteristics (novelty, autonomy, imminence).

NatSec Incident View

National Security impact assessment for each AI incident in the database, including 5 categories of impact and threat classification (novelty, imminence, autonomy)

NatSec Impact

View how the National Security impact of incidents has changed over time in each of the impact and threat classification categories.

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form.