Claude shows language‑dependent brand recommendations in travel queries
The author notes that many users in non‑English speaking regions, such as Persian‑speaking residents of Los Angeles County, are unlikely to switch to English when interacting with AI.
Key points
- The author’s 1,200‑response test used Claude Sonnet 5 across five languages
- Anthropic’s July study sampled 310,000 conversations across 20 languages
Anthropic published a July report that sampled just under 310,000 real conversations across the 20 most common languages, split between three models (Sonnet 4.6, Opus 4.6, Opus 4.7). The study found that Claude’s values shift with language, showing greater warmth in Hindi and Arabic and greater rigor in English and Russian. The author ran his own tests on six shopping or booking scenarios in five languages (English, Arabic, Russian, Simplified Chinese, Persian) using Claude Sonnet 5 with web search off.
Results revealed large language‑dependent swings in brand recommendations. In the airline scenario, Claude chose British Airways 52.5 percent of the time in Persian but never in Russian, where it favored easyJet 72.5 percent of the time. In Chinese, Bahía del Duque was selected 43.6 percent of the time versus 7.7 percent in Arabic. The author notes that such disparities could affect travel and tourism marketers who test only one language.
The story so far
5 episodes →- Claude shows language‑dependent brand recommendations in travel queriesthis story
Lost In Translation: How Language Changes Which Brand AI Recommends
Search Engine Journal · 13 October 2026
Loading the full article…
This text was published by Search Engine Journal and written by Reza Moaiandin. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Marketing & Small Business
All →- Character.AI files for Chapter 11 bankruptcy after five years · 7 src
- HubSpot AEO helps marketers target AI answer engine citations · 12 src
- Honeyjar launches AI tools for PR and comms workflows · 1 src
- Marketers report over 90% AI use but struggle to integrate tools into workflows · 1 src
- Ahrefs studies 15M ChatGPT fan-out queries to map search behavior · 1 src
Comments
via GitHub Discussions