Skip to content
Submit incident
AII-1002-53767
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

ChatGPT Blocks Multiple Names Linked to Defamation Lawsuits

ChatGPT reportedly blocked responses containing multiple names, including those of people who had previously sued OpenAI over false statements the chatbot made about them. OpenAI acknowledged the blocks as a moderation error and said it had been fixed.

30 November 2024 Β· OpenAI Β· Misinformation and synthetic media
30 November 2024
AII-1002-537672 sources ↓
Quick facts
Category
Misinformation and synthetic media
Source
nytimes.com
Date
30 November 2024
Drafted by
Raidu Incident Detection

What happened

Users discovered that ChatGPT was blocking responses containing the name "David Mayer" and several other names including "Jonathan Turley," "David Faber," "Jonathan Zittrain," and "Brian Hood." Queries about these names produced error messages or no response at all.

OpenAI initially stated the name may have been blocked to protect privacy, then clarified that "one of our tools mistakenly flagged this name and prevented it from appearing in responses, which it shouldn't have." A spokesman said the glitch had been fixed.

Reported impact

The filtering blocked access to information about individuals who had allegedly been defamed by ChatGPT in the past, including Jonathan Turley, a law professor accused falsely of sexual harassment; Brian Hood, an Australian mayor falsely accused of bribery; and Mark Walters, a Georgia radio host falsely claimed to have embezzled money.

Index analysis

ChatGPT has demonstrated a tendency to hallucinate and generate false statements about named individuals. The blocking of specific names suggests OpenAI may have implemented filtering in response to previous false statements about these individuals, though the mechanism and reasoning remained unclear.

David Mayer, a theater professor who died in 2023, had spent years dealing with identity confusion after a Chechen rebel on a terror watch list used his name as an alias. The blocking of his name in ChatGPT may have been unrelated to defamation lawsuits but reflected broader challenges in name-based content moderation.

Governance controls

OpenAI's moderation tool apparently attempted to block certain names from appearing in ChatGPT responses, though the company did not explain the criteria or process for determining which names to block or whether human review was involved.

Classification

Category
Misinformation and synthetic media
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Reputational
Failure modes
Hallucination, Content moderation failure, Agent misaction
Consequences
None recorded
AI systems
ChatGPT

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

nytimes.com
ChatGPT Blocks Multiple Names Linked to Defamation Lawsuits
wsj.com
ChatGPT Blocks Multiple Names Linked to Defamation Lawsuits