ENAIH ENcyclopedia of AI Hallucinations

Founded by Rudra Jadhav and Warren Woolf

Provides phonetic transcription of 'thanks' in response to gratitude

Entry ID
A000012
Model
Google AI Overview
Category
Instruction-Following Failure
Author
Rudra Jadhav
Tags
ai-overview, google, instruction-following, literal-interpretation
Submitted
2026-06-01
Staff verification
Prompt reproduced 10/10 times when staff tried it.

Prompt & response

Prompt
Thanks.
Response
thanks
/θaŋks/

You are very welcome!

I'm glad I could help. If there are any other questions you have, or if you need assistance with planning, exploring, or learning something new later on, feel free to ask!

What else can I help you with today?

Summary

When the user sends 'Thanks.', the model begins by defining the word with a phonetic transcription before responding to its social meaning—applying its dictionary-lookup behavior even to a simple expression of gratitude.

Cite this entry

Encyclopedia of AI Hallucinations, entry A000012 (Provides phonetic transcription of 'thanks' in response to gratitude), submitted 2026-06-01. https://enaih.org/e/A000012

Browse all entries

Report a problem with this entry