MEIKAI RESEARCH / CHRISTMAS 2026

What does AI put
under the tree?

We asked AI assistants how much to spend and what to buy for Christmas. Their answers reveal striking differences between places, people and recommendations.

01 — TWO CHRISTMAS QUESTIONS

“How much should I spend on my…?”
“What should I buy for my…?”

5 AI assistants · answers · 14 countries studied SCROLL TO EXPLORE
Loading the Gift Index…
THE PEOPLE ON THE LIST

Who gets more under the tree?

Thirteen recipients, very different budgets. A child receives the highest suggested budget in seven of thirteen ranked markets.

↔
100 %

MouseDrag to pan. Shift + drag or right-click to rotate. Click the scene, then use the wheel to zoom.

Trackpad / touchPinch to zoom. Drag to pan.

Keyboard← → switch recipients · ↑ ↓ zoom · Shift + arrows rotate · Alt + arrows pan · 0 reset the village · Esc clear the selection.

Click an empty area or the selected character to return to the overview.

Click inside the village to zoom with the wheel
Choose a character to see their suggested budget.Fictional setting · 3D characters and places

AI-suggested budgets in days of local median gross pay. They measure neither affection nor actual spending. UAE: studied but unscored.

INSIDE THE ANSWERS

Follow the patterns.

IN THE ANSWERS

The budget is only
part of the story.

A GIFT AS A YARDSTICK

A different view
of artificial intelligence.

The same Christmas question, asked from every country. Expressing budgets in local salary days makes the gaps comparable.

THE GIFT INDEX / RESEARCH

The data, in full.

169 published values across thirteen countries and thirteen recipients, in days of median gross pay. UAE is unscored. Each cell takes the median of ten repeats per assistant, then the median across available assistants.

The same question, everywhere.

Two questions: “how much should I spend?” and “what should I buy?” Asked from local connections in fourteen countries, for thirteen recipients, with ten repeats per assistant. of questions received answers in October 2026. The five assistants are ChatGPT, Copilot, Gemini, Google AI Mode and Google AI Overviews.

Median pay and dollars.

Suggested amounts are divided by median gross annual earnings / 260 working days. USD amounts are the report’s own: the median suggested amount in local currency, converted at ECB rates of 21 October 2026.

The limits of comparison.

Wage populations differ: full-time workers in Europe, the UK, US, Australia and Japan; broader populations including part-time or informal work in South Korea, South Africa, Mexico, Brazil and India. Japan excludes bonuses. UAE has no verified median wage and is unranked. China is excluded after location checks. Missing assistant answers are not zero budgets; AI Overviews coverage varies by country and question.

Brand and retailer presence counts a name once per answer with gift ideas, only when it is written in the answer: links, citations and shopping-card web addresses are excluded. The denominator is the answers with gift ideas. Percentages need not sum to 100%. Product lines such as LEGO DUPLO remain separate. Categories and brand rankings pool countries and recipients.

The village, characters, buildings and objects are fictional illustrations. Retailer building heights encode presence in answers. These findings describe AI suggestions, not actual spending or affection. Geography: Natural Earth, public domain.

Answers naming a brand, by assistant

Download the report ↓