The Responsible AI Collaborative (TheCollab) is a nonprofit that maintains the AI Incident Database (AIID), the leading public repository of documented real-world AI harms and near-harms.
The Responsible AI Collaborative (TheCollab) is a nonprofit that maintains the AI Incident Database (AIID), the leading public repository of documented real-world AI harms and near-harms.
People
Updated 05/18/26Artificial Intelligence Incident Database Consulting Editor
Funding Details
Updated 05/18/26- Annual Budget
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
Org Details
Updated 05/18/26The Responsible AI Collaborative (TheCollab) is a nonprofit organization founded in 2022 and headquartered in Los Angeles, California. It was formed as a special-purpose independent entity to govern and sustain the AI Incident Database (AIID), which had originally been launched by the Partnership on AI in November 2020. The AIID is the most widely used public repository of documented real-world AI harms. It contains over 5,800 human-annotated incident reports covering cases where AI systems caused harm or near-harm to individuals and society. The database has informed the development of national and intergovernmental AI standards and is used by central banks, regulators, researchers, and practitioners worldwide. TheCollab was co-founded by Sean McGregor, a machine learning safety researcher who started the AIID and previously worked with the Digital Safety Research Institute at UL Research Institutes. The organization operates with a small staff of roughly four full-time employees, supplemented by volunteer editors and contributors from partner organizations. Leadership has included Scott Allen Cambo and Neama Dadkhahnikoo as executive directors. The organization is structured as a collaborative to reflect the participatory nature of its work: it relies on incident editors, taxonomy editors, open-source contributors, and organizational partners to maintain and expand the database. In 2023, Open Philanthropy awarded TheCollab a $100,000 grant to support the AIID. The organization also has a corporate sponsor cohort, with Trustible leading the inaugural group announced in 2025. TheCollab's work directly supports AI risk management and governance by providing a structured historical record of AI failures that practitioners and policymakers can draw on when designing safer systems and policies.
Theory of Change
Updated 05/18/26TheCollab believes that systematically cataloging real-world AI harms enables the AI development community, regulators, and the public to learn from past failures in the same way that aviation and cybersecurity communities use incident databases to improve safety. By making this incident data freely available and well-structured, the organization aims to: (1) help developers identify and avoid patterns of failure in deployed AI systems; (2) inform risk frameworks, standards, and regulation with empirical evidence; and (3) create accountability pressure that incentivizes more careful AI deployment. The causal chain runs from incident documentation to improved situational awareness for stakeholders, to better risk management practices, to reduced probability and severity of AI-related harms.
Grants Received
Updated 05/18/26Projects
Updated 05/18/26A free, open-source database maintained by the Responsible AI Collaborative that indexes real-world harms and near-harms from deployed AI systems so people can learn from past failures.
Discussion
No comments yet. Be the first to share your thoughts.