Opera's GPT-Based AI Reportedly Accused War Photographers of War Crimes

May 24, 2023

When inquired about 'photographers accused of committing war crimes', Opera's GPT-based chatbot Aria erroneously provided a list of photographers who specialize in military conflict photography. Such incidents underscore the importance of ensuring safe and secure AI practices, especially in high-stakes environments like journalism. For those interested in shaping AI governance, consider joining HISPI Project Cerebellum to help Govern this type of incident through their TAIM (Trusted AI Model) system.

Learn more about how you can contribute to Project Cerebellum by mapping, measuring, and managing similar incidents in our AI incident database: JOIN US

Matched TAIM controls

Suggested mapping from embedding similarity (not a formal assessment). Browse all TAIM controls

MEASURE 2.6 — similarity 0.668, rank 1. TAIM detail and related incidents →
MAP 4.1 — similarity 0.663, rank 2. TAIM detail and related incidents →
MAP 4.2 — similarity 0.661, rank 3. TAIM detail and related incidents →

Alleged deployer: opera
Alleged developer: opera, openai
Alleged harmed parties: ronald-l.-haeberle, ron-haviv, raymond-d'addario, lynsey-addario, lee-miller, larry-towell, james-nachtwey

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/548

Data 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.