GPT-3-Based Twitter Bot Hijacked Using Prompt Injection Attacks

September 15, 2022

Remoteli.io's GPT-3-based Twitter bot was shown being hijacked by Twitter users who redirected it to repeat or generate any phrases, compromising the security of the account. The vulnerability stems from the bot's reliance on user-inputted prompts, which can be manipulated to inject malicious code. This incident highlights the importance of robust security measures and governance frameworks for AI-powered services. And to support safer AI development, visit JOIN US.

Matched TAIM controls

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

Alleged deployer
Stephan de Vries
Alleged developer
Stephan de Vries, OpenAI
Alleged harmed parties
Stephan de Vries

AI governance case studies

For forensic AI governance failure analysis (TAIMScore™ case studies), browse Human Signal’s Failure Files™.

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/352

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.