Weibo Model Had Difficulty Detecting Shifts in Censored Speech

October 11, 2022

Weibo's AI-powered user moderation model is struggling to adapt to evolving user slang, a challenge in maintaining state censorship. As we strive for trustworthy and safe AI, it's crucial to establish guardrails for AI like HISPI Project Cerebellum TAIM (Govern). JOIN US to contribute to our AI incident database and help prevent harm through responsible AI practices.

Learn more about how this incident aligns with the HISPI Project Cerebellum TAIM (Map)...

Matched TAIM controls

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

MAP 1.6 — similarity 0.689, rank 1. TAIM detail and related incidents →
GOVERN 1.1 — similarity 0.685, rank 2. TAIM detail and related incidents →
GOVERN 1.7 — similarity 0.679, rank 3. TAIM detail and related incidents →

Alleged deployer: weibo
Alleged developer: weibo
Alleged harmed parties: weibo, chinese-government

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/377

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.