The AIID is largely developed by the partnership community, but development of both the data product and the applications built on top of the data is open to contributors. Regular contributors to the database will be invited, but not expected, to take leadership roles in the system's development.
Phase 1: Initial Incident Collection
The initial dataset was collected in 2019 by combining incident listings from Roman Yampolskiy, Catherine Olsson, and Sam Yoon. These incidents are associated with more than 1k total incident reports.
Phase 2: Incident Discovery Application
The user stories supported in this phase include,
- system architects/product leads: "I want to find instances where similar systems have failed in the real world so that I can proactively avoid those failures in my system design."
- public relations managers: "I want to see what has happened in the past so I can understand how to avoid embarassment in the future."
- lawyers: "I want to discover incident history so I can understand how to avoid liability in the future."
Phase 3: Flexible Taxonomy (Current)
Intelligent systems are in development for every market segment and function of government. Developing the research quality of the data product requires systems for regularizing and classifying incidents according to,
- Incident types
- Incident scale
- Technologies involved
- Impacted parties
These, and other, classifications are a point of contention in multi-stakeholder systems. To advance the research utility of the database while not requiring full consensus on every classification, it is possible to apply taxonomies independently of a central taxonomic authority. Each "taxonomic scope" is managed by an editor that develops the taxonomy, and is responsible for applying the taxonomy across incidents. Contact us for more information.
Phase 4: Incident Monitoring
As more intelligent systems are deployed within the real world, it will become increasingly difficult to monitor, collect, and categorize incidents. A scraper that monitors content sources for AI incidents and facilitates the easy ingestion thereof would ensure the AIID continues to be relevant for the most recently developed intelligent systems.