Skip to content
Submit incident
AII-1001-87309
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

Arkansas teacher suspended after AI generated test question with racial slur

An AI tool approved by Fort Smith School District in Arkansas generated test questions containing a racial slur when an English teacher asked it to create comprehension questions for a lesson on a memoir about school desegregation. The Fort Smith School Board suspended the teacher for 90 days without pay after the incident.

1 April 2025 Β· Robert Taylor Β· Harmful and abusive content
1 April 2025
AII-1001-87309
Quick facts
Category
Harmful and abusive content
Country
United States
Date
1 April 2025
Drafted by
Raidu Incident Detection

What happened

Robert Taylor, an English teacher at Northside High School in Fort Smith, Arkansas, used an AI tool called Magic AI approved by the district to generate comprehension questions for a lesson on 'Warriors Don't Cry,' a memoir by Dr. Melba Patillo Beals about her experiences as a Black student during school desegregation. The AI-generated test questions included a racial slur.

Reported impact

The test questions containing the racial slur were administered to students. The original district recommendation was to terminate Taylor from his position.

Index analysis

Taylor stated he directed the AI tool to write comprehension questions about literary devices only and did not ask it to write about ethical, logical, or emotional use of specific words. He contended the slur appeared in the original book to demonstrate how harmful it was to the author during her experience.

Governance controls

The Fort Smith School Board voted 6-to-1 to suspend Taylor without pay for 90 days effective immediately, modifying the district's original recommendation to terminate him after hours of discussion at a Monday night meeting.

Classification

Category
Harmful and abusive content
Industry
Education and training
Country
United States
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Human or fundamental rights, Psychological
Harmed
Children
Failure modes
Hallucination, Content moderation failure
AI systems
Magic AI

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident