AI Incident Tracker

NatSec Incident View

View detailed assessments of the National Security impact of each incident in the AI Incident Database, including 5 categories of impact and threat classification (novelty, imminence, autonomy) based on this National Security Impact Framework.

Complex queries can be built by adding filters across multiple categories.

‍

Preset filter configurations:

Important context for interpreting these results:

The data presented is the output of an LLM classifier pipeline applied to the raw reports from the AI Incident Database (AIID), which relies on submissions from the public and subject matter experts. The quality, reliability and depth of detail in the reports varies across the dataset. As the reporting is voluntary, the dataset is inevitably subject to some degree of sampling bias.
The LLM classification tool has been developed iteratively, and its agreement with human expert consensus is comparable to the agreement between 2 independent human experts (article forthcoming). Spot-checks have been used to provide feedback on misclassifications and to iterate the tool, improving its reliability, however there are likely to remain incidents where the LLM classification does not match expert consensus.

Therefore patterns and trends observed in the data should be taken as indicative and validated through further analysis.

‍

Explore Other Views

Incident Timeline

View numbers of reported incidents each year split or filtered by risk domain, entity, timing, harm severity or primary AI purpose.

Incident Overview

View 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. Filter by risk domain, timing, entity, AI purpose or severity of harm caused.

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.

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

NatSec Impact Framework

This framework offers a structured methodology to assess the national security implications of AI-related incidents. It is intended to be applicable across different nations, grading the impact to the primary target country and identifying other nations that were also affected.

AI Risk Repository

The AI Risk Initiative helps organisations understand and prioritise emerging AI risks, and learn how leaders are responding.

RisksIncidents MitigationsGovernanceBlogTeam

© MIT FutureTech 2025
The MIT AI Risk Initiative is licensed under CC BY 4.0
MIT Accessibility