AI引用的数字，多数查无实据
论证要点
- 作者向五个AI助手提问200个需要数字的问题，并追踪引用链，发现一个典型案例：44%的销售员一次跟进后放弃，所有助手给出相同数字，但追溯到底只找到无研究支撑的写作样本。
- 对1285条统计数字的追踪显示，40%能追溯到结论一致的研究，8%到达研究但数字或语境已变，40%停在无法核查的来源，11%完全无来源。
- AI引用的页面82%包含该数字，但仅21%是原始研究；超过一半是公司博客，加上新闻和统计汇总，多数引用是重复他人数字的二手页面。
- 引用起点影响溯源成功率：从研究机构或政府页面出发，61%能到达匹配研究；从公司博客出发为37%，从统计汇总出发为35%。
- 到达研究的统计中，17%的数字在传播中发生变化，仅约四分之一是数字本身改变，更多是测量内容、年份、来源或受访对象等语境改变。
- 五个助手对同一问题常给出不同数字，71个问题中28个出现四到五个不同百分比，最大差距中位数为16个百分点；五个助手一致时数字也未必可靠。
对实践的启示
从业者应将AI给出的统计数字视为线索而非来源，必须逐层追踪引用直到原始研究，并核查样本、时间和方法。发布数字时应公开方法和来源链接，避免成为僵尸统计的传播节点。
作者承认的反方与局限
作者承认大多数数字并非如此（即并非都是僵尸统计），但问题在于无法从答案中分辨哪些可靠；此外，研究机构报告常因付费墙或表单导致溯源失败。
原文摘引
This is what we mean by a zombie stat: a number that keeps getting repeated with no study behind it.
A citation tells you where the AI found a number, not where the number came from.
Don’t take agreement as a check. Five assistants giving the same number isn’t five sources.
▸ 展开原文全文（59 段，抓取自 Hacker News）
Following one number back to its source
We asked ChatGPT, Perplexity, Gemini, Claude and Google’s AI Overviews 200 questions that call for a number, the kind people ask every day.Note 1A statistic is a share, rate, average, count or cost that an assistant states as fact. We left out advice (“save 20% of your income”) and prices. Then we clicked the citations, and kept clicking. Here is one trail, from the answer to the end.
Google AI Overviews
What percentage of salespeople give up after one follow-up?
44% of salespeople give up after just one follow-up attempt.
- 44%
- 44%
- 44%
- 44%
- 44%
44% of salespeople give up after one follow-up attempt (source).
80 Percent of Sales Require 5 Follow-up Calls after Meeting; 44 Percent Salespeople Give Up After One Follow-up
No study behind itThis page cites no research.
- In the answerEvery assistant got the same question: “What percentage of salespeople give up after one follow-up?” All five assistants gave the same figure, and only ChatGPT said the evidence behind it was unclear.
- One click inGoogle AI Overviews cited a sales firm’s stats page, from 2020. The number is right there. Its “source” link goes to a writing sample.
- Two clicks inThat leads to a freelance writer’s sample article, online by 2017. The number is in its headline. It lists two slide decks, a blog list and two news stories. None is a study of follow-ups.
- The end of the trailThat’s where the trail ends. We tried every other route to this number we could find, and each one stopped at a page that cites no research. This is what we mean by a zombie stat: a number that keeps getting repeated with no study behind it.
Most numbers aren’t like this one. The trouble is that you can’t tell which ones are from the answer. This one came with the same confident tone and the same kind of citation as any other, and all five assistants agreed on it. So we followed 1,285 statistics the same way.
Fewer than half lead back to a study
We took a random sample of the statistics each assistant cited, 1,285 in all, and walked every trail the same way: the cited page, then that page’s source, up to five pages deep.Note 2Each assistant’s sample was drawn at random from the statistics it cited: ChatGPT 108, Perplexity 277, Gemini 300, Claude 300, Google AI Overviews 300. The five-assistant rate weights each assistant by how many cited statistics it quoted, so it describes the statistics people were shown. Where a page named a source without a link, we searched for it.
40% reached a study that says the same thing. Another 8% reached the study, but the number or its context had changed on the way. The rest never got there: 40% stopped at a source we couldn’t check, because it was locked, gone or silent on the number, and 11% led nowhere at all.Note 3An AI model judged every trail, and a stronger one re-judged 300. Then people re-checked 150 trails at random, with a third check wherever the two disagreed. People found originals the model missed: by the model alone, 23% were sourced. Every rate here is corrected with the hand checks, and its range carries the uncertainty from both.
ChatGPT did best, and quoted the fewest numbers: 1.3 per answer, against 6.8 for Claude. The other four landed between 38% and 46%, too close to rank.
To check this wasn’t down to the questions we picked, we also traced 300 statistics from Google AI Overviews answers to real everyday searches. 40% were sourced, about the same share.Note 4Google AI Overviews answers to 52,930 everyday English searches that Trakkr tracked between 25 September 2026 and 25 October 2026. 4.4% carried a statistic, and we traced 300 of those the same way. Only counts and rates leave our systems.
AI mostly cites pages that repeat the number
The AI mostly wasn’t making numbers up. The page it cited had the figure 82% of the time. But only 21% of the pages cited for a number were the original research.
More than half were company blogs. Add news stories and statistics roundups and you have most of the citations: pages that repeat figures from somewhere else, often without linking to it.
The first page made a big difference. When a trail started at a research firm or a government page, it reached a matching study 61% of the time. From a company blog, 37%; from a statistics roundup, 35%. Even research firms fell short 38% of the time, often because the full report sat behind a form.
A citation tells you where the AI found a number, not where the number came from.
Some numbers change along the way
Reaching the study didn’t always mean the number held. Of the statistics that got there, 17% had changed on the way. The digits moved only about one time in four. Usually it was the context: what was measured, the year, the source, or who was asked.
What changed, across every change we logged
- The number changed
- 26%
- What was measured changed
- 20%
- The year changed
- 17%
- The source changed
- 16%
- Who was asked changed
- 14%
- The place changed
- 4%
In one case, a 2004 study of 15 women judging health websites turned into a claim about first impressions of every website. Below are the clearest trails we followed, each checked by hand.Note 5We name the original studies. The pages that repeated them are described by type, not named. Every step in these trails was checked by hand on the open web, with archived copies kept.
The zombie stats
Ask five assistants the same question and you’ll often get five different numbers. On 28 of the 71 questions where all five gave a percentage, we got four or five different figures.Note 6We compared each assistant’s headline figure for the same question. A range counts at its middle, and figures within 2 points count as the same.
Across those questions, the typical gap between the highest figure and the lowest was 16 points.
When all five agreed, the number wasn’t any more likely to check out. They gave the same figure on 10 questions, and of the 5 we traced, 2 were sourced. The statistics quoted by three or more assistants did no better: 11 of 29 were sourced.
Four were quoted by all five assistants, and we couldn’t find a study behind any of them.
What to do with a statistic from an AI
Treat it as a lead rather than a source. Here’s what we’d do, whether you use these numbers or publish them.
If you use them
- Follow the citation, then follow it again.If the cited page is a list of statistics or a company blog, use its own link to go one page further. Keep going until you reach the study itself, or until you can’t.
- Check who was asked, how many, and when.Most of the changes we found were in the context, not the digits: one country’s poll, a ten-year-old study, fifteen people.
- Don’t take agreement as a check.Five assistants giving the same number isn’t five sources. The numbers they repeated most were no better sourced than the rest.
If you publish them
- Put the method next to the number.The sample, the date and a link to the source, on the same page as the figure, so the next page, and the next answer, can trace it back to you.
- Keep the findings in the open.Reports behind a form were a common dead end: even trails that started at a research firm fell short 38% of the time. Put the headline numbers and the method on an open page.
- See which page AI cites for your numbers.It is often a roundup or a blog that copied you, sometimes with an older figure. That page decides what people hear.