Incident #399·Misinformation

Meta AI's Scientific Paper Generator Reportedly Produced Inaccurate and Harmful Content

November 15, 20223 reportsView on AIID

Description

Meta AI trained and hosted a scientific paper generator that sometimes produced bad science and prohibited queries on topics and groups that are likely to produce offensive or harmful content.

Actors

Alleged Deployers

meta
facebook

Alleged Developers

meta
facebook

Harmed Parties

minority-groups
meta
facebook
minority-groups

Classification

Risk Domain

Misinformation

Risk Subdomain

3.1. False or misleading information

MIT Risk Profile

AIUnintentionalPost-deployment

Source

Databaseaiid

LicenseCC BY-SA 4.0

View on AIID

Related Incidents (Misinformation)

View all →

#541

ChatGPT Reportedly Produced False Court Case Law Presented by Legal Counsel in Court

May 2023

#616

Sports Illustrated Is Alleged to Have Used AI to Invent Fake Authors and Their Articles

Nov 2023

#645

Seeming Pattern of Gemini Bias and Sociotechnical Training Failures Harm Google's Reputation

Research Papers on This Topic

Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature

450 citationsOct 2023

Why Language Models Hallucinate

371 citationsSep 2025

CIFAKE: Image Classification and Explainable Identification of AI-Generated Synthetic Images

283 citationsMar 2023

Degenerate Feedback Loops in Recommender Systems

265 citationsFeb 2019

Source Reports(3 total)

Thread by @osanseviero on Thread Reader App

threadreaderapp.com· Nov 2022

New Meta AI demo writes racist and inaccurate scientific literature, gets pulled

arstechnica.com· Nov 2022