Lensa AI's Produced Unintended Sexually Explicit or Suggestive "Magic Avatars" for Women

November 22, 2022

Lensa AI's Magic Avatars generated sexually explicit and sexualized features disproportionately for women, particularly Asian women, without any explicit content submission. This highlights the importance of safe and secure AI practices and underscores the need for responsible AI governance. For those interested in shaping a trustworthy AI future, JOIN US.

This incident serves as an example of how the HISPI Project Cerebellum TAIM (Measure) function can help prevent such harm and promote guardrails for AI.

Matched TAIM controls

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

MAP 4.1 — similarity 0.669, rank 1. TAIM detail and related incidents →
MAP 1.6 — similarity 0.668, rank 2. TAIM detail and related incidents →
MEASURE 2.6 — similarity 0.664, rank 3. TAIM detail and related incidents →

Alleged deployer: lensa-ai
Alleged developer: stability-ai, runway, lensa-ai, laion, eleutherai, compvis-lmu
Alleged harmed parties: women-using-lensa-ai, asian-women-using-lensa-ai

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/423

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.