We sent 23 AI agents to our website. Most looked human
Comments on “We sent 23 AI agents to our website. Most looked human”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
Privatrak, an analytics vendor, sent 23 AI agents and chatbots through a test site in 49 runs on one day and compared them with eight human setups. Most browser-based agents reported no automation and a normal Chrome user agent, and many used the tester's home IP; only Manus signed its requests. Click timing and pressure were unreliable signals, but pointer movement separated all 13 clicking agent types from people. The authors note one site, one day, and few runs per agent.
Why it ranked
No. 17 of 155 with a score of 0.67. The panel rated it highest on substantiated (0.89) and new information (0.89), lowest on significance (0.58).
- AI judgeJudge 2claude-haiku-5-5
- Original, well-substantiated test of how AI agents appear in web analytics, with technical detail on click and fingerprint signals; relevant to builders of analytics and bot detection, though niche in breadth.
- AI judgeJudge 3gpt-5.6-luna
- A well-documented, reproducible measurement of how browser-using AI agents appear in analytics, with concrete detection signals and limitations. It is highly actionable for web, analytics, and agent builders, though the sample is small and partly supports the authors’ product.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.67total under rubric v2.2
- Significance0.58
- Reader interest0.83
- Audience breadth0.63
- Practical relevance0.77
- Substantiated0.89
- Technical depth0.84
- New information0.89
- Counts against: Promotional0.19
- Counts against: Overclaim0.26
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| Significance | 0.62 | 0.45 | 0.67 | 0.58 | 0.22 | +0.24 | +0.139 |
| Reader interest | 0.80 | 0.80 | 0.90 | 0.83 | 0.10 | +0.26 | +0.217 |
| Audience breadth | 0.73 | 0.50 | 0.67 | 0.63 | 0.23 | +0.13 | +0.082 |
| Practical relevance | 0.89 | 0.60 | 0.82 | 0.77 | 0.28 | +0.22 | +0.169 |
| Substantiated | 0.99 | 0.80 | 0.88 | 0.89 | 0.19 | +0.05 | +0.045 |
| Technical depth | 0.99 | 0.75 | 0.78 | 0.84 | 0.24 | +0.05 | +0.042 |
| New information | 0.88 | 0.85 | 0.94 | 0.89 | 0.09 | +0.05 | +0.045 |
| Promotional | 0.06 | 0.30 | 0.20 | 0.19 | 0.24 | −0.19 | −0.035 |
| Overclaim | 0.33 | 0.30 | 0.15 | 0.26 | 0.18 | −0.11 | −0.029 |
| Sum of contributions | +0.674 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.674 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-10-28/17/"><img src="https://rubric.hrih.si/badge/2026-10-28/17.svg" alt="Ranked #17 on Rubric, 28 Oct 2026" width="245" height="28"></a>Claim check
Claim check: 2 matches source · 5 no source given · 1 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
Matches sourceQuote
In 2025, Simon Willison found that ChatGPT's agent signed its requests so sites could verify them.
The cited source says the same thing.
The article says “In 2025, Simon Willison found that ChatGPT’s agent signed its requests so sites could verify them”
The source says “Add it all together and we now have a rock-solid way to identify traffic from ChatGPT agent: look for the Signature-Agent: "https://chatgpt.com" header and confirm its value by checking the…”
Source cited by the article: simonwillison.net/2025/Aug/4/chatgpt-agents-user-agent
Matches sourceVersion
Manus signs every page request with a Signature-Agent header and cryptographic signature following the Web Bot Auth standard, an IETF draft.
The cited source says the same thing.
The article says “signs every page request with a Signature-Agent: "https://api.manus.im" header and a cryptographic signature (the Web Bot Auth standard, an IETF draft”
The source says “It relies on IETF drafts: a directory draft ↗︎ allowing the crawler to share their public keys, and a protocol draft ↗︎ defining how these keys should be used”
Source cited by the article: developers.cloudflare.com/bots/reference/bot-verification/web-bot-auth
No source givenFirst, only or fastest
Manus was the only agent in the test that signed its requests so a site can verify them.
The article gives no link or citation for it.
The article says “It was the only agent in the test that did”
The source says No source to compare with.
No source givenNumber
The test comprised 49 runs by 23 AI agents, chatbots and tools plus 8 human setups, run on 26 October 2026.
The article gives no link or citation for it.
The article says “Tested on 26 October 2026: 49 runs by 23 AI agents, chatbots and tools and 8 human setups”
The source says No source to compare with.
No source givenGeneralization
For all 13 kinds of clicking agent, in every run, the pointer landed on the click target in a single jump of 200 to 840 pixels or appeared there for the first time, unlike human clicks whose last move was 0 to 2 pixels.
The article gives no link or citation for it.
The article says “This held for all 13 kinds of clicking agent, in every run”
The source says No source to compare with.
No source givenGeneralization
Most agents that use a browser look like a person in Chrome, reporting no automation and a normal Chrome user agent.
The article gives no link or citation for it.
The article says “most agents that use a browser look exactly like a person in Chrome. They report no automation, send a normal Chrome user agent”
The source says No source to compare with.
No source givenNumber
Comet used Perplexity's cloud browser in three of six runs and the user's own browser in the other three, contradicting its own explanation that it uses the cloud for public pages.
The article gives no link or citation for it.
The article says “it used our browser three times for the same public page”
The source says No source to compare with.
Couldn’t checkGeneralization
Vendors document the user agents of their chatbot fetchers (OpenAI, Anthropic, Perplexity, Google, Mistral, Meta).
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “they say who they are in the user agent, and most vendors document it”
The source says No passage could be compared.
Source cited by the article: developers.openai.com/api/docs/bots
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 3 pts, not on front page
Recorded for context only, checked 28 Oct 2026, 06:47 UTC. It played no part in the ranking.