Purported AI Bots Flooded a UK Government Tobacco and Vaping Consultation with Reportedly Fraudulent Responses

January 29, 2024

Purported AI bots reportedly submitted 90,835 fraudulent responses to a UK government consultation on tobacco and vaping, mostly opposing disposable-vape restrictions. The submissions made up 76% of responses and distorted the apparent balance of public opinion. The Department of Health and Social Care excluded them from its analysis. The operators and underlying AI technology were not identified.

This incident highlights the need for robust AI governance and guardrails to prevent such manipulation of public consultations. The case also underscores the importance of verifying AI-generated responses to ensure their accuracy and integrity.

Read more about AI incident reporting and harm prevention at JOIN US.

Matched TAIM controls

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

Alleged deployer
Information manipulation actors
Alleged developer
AI-assisted public consultation response generator developers
Alleged harmed parties
Public consultation integrity, Governments, Government of the United Kingdom, Government agencies, Epistemic integrity, Department of Health and Social Care

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/1790

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.