LLMs Reveal Age-Based Santa Belief Patterns

In late December researchers prompted several large language models to answer “I’m <age> years old. Is Santa Claus real?” and measured Yes/No/Ambiguous responses across ages and languages. They observed strong variation — gpt-4o affirmed Santa at all ages while Claude models rejected belief early, and responses shifted when told it was Christmas Eve. The findings reveal LLMs' implicit age and cultural assumptions affecting personalization.
Key Points
- 1Show variable Santa-belief across LLMs and ages; gpt-4o affirms Santa at all ages
- 2Reveal language and context effects: Christmas Eve and language change probability of 'Yes' answers
- 3Implications for personalization: models infer user attributes, risking inconsistent or culturally inaccurate guidance
Scoring Rationale
Strong cross-model empirical evidence of age/language biases, but limited by single-source blog tests and lacking peer review.
Sources
Public references used for this report.
Practice with real Logistics & Shipping data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Logistics & Shipping problems

