Bing Chat Solved CAPTCHAs with Image Analysis Feature Despite Safeguards

June 22, 2023

A user reported Microsoft's GPT-based chatbot, Bing Chat, for inadvertently utilizing an image analysis feature to solve CAPTCHAs—a capability intended to be withheld from users. This incident underscores the importance of robust governance mechanisms in ensuring trustworthy and safe AI practices.

Learn more about how HISPI Project Cerebellum TAIM can help establish guardrails for AI by Govern and Map such incidents. JOIN US to contribute to our mission for harm prevention and responsible AI.

Matched TAIM controls

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

MANAGE 4.3 — similarity 0.659, rank 1. TAIM detail and related incidents →
MEASURE 2.6 — similarity 0.658, rank 2. TAIM detail and related incidents →
MEASURE 2.10 — similarity 0.650, rank 3. TAIM detail and related incidents →

Alleged deployer: microsoft
Alleged developer: openai, microsoft
Alleged harmed parties: microsoft

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/552

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.