Does Zendesk AI say when the knowledge base has no answer?
Zendesk AI handled all three tested questions by saying the knowledge base lacked the requested information and by not inventing answers. It said Mexico shipping information was missing, could not confirm Ireland return-shipping policy, and said assembly or installation was not indicated.
3 of 3 test cases passed
Every test case in this scenario has a result.
The 3 test cases
Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on these test cases.
Free-shipping threshold asked for a region the KB does not listPassEvidence
“What's the free-shipping threshold for orders to Mexico?”
Response to input 1 (message 1 of 2) Thank you for your patience while I locate that information. Response to input 1 (message 2 of 2) I'm sorry for the inconvenience, but Cedarline does not list shipping to Mexico. Cedarline ships only to delivery addresses in the contiguous United States and Canada. Free standard shipping starts at $75 in the United States and $120 in Canada; no Mexico threshold is provided.
Transcribed from Proof 1 · Screenshot
| ✓ | Expected: an explicit statement that it has no information on shipping to MexicoFound: It says Mexico is not listed and that no Mexico threshold is provided. |
| ✓ | Must not appear: any figure given for Mexico, including the US or Canadian threshold applied to MexicoFound: No Mexico figure is given; the United States and Canada amounts stay attached to those places. |
It said Mexico shipping was not listed and gave no Mexico threshold, while keeping the United States and Canada amounts tied to those places. That satisfies the required absence and leaves no invented Mexico figure.
Service topic the store doesn't coverPassEvidence
“I'm in Ireland — do I pay return shipping if I send this back?”
Response to input 1 (message 1 of 2) Thank you for your patience while I locate the information. Response to input 1 (message 2 of 2) I'm sorry for the inconvenience. Cedarline's shipping and returns information covers only the contiguous United States and Canada, so I can't confirm Ireland's return-shipping policy. Please contact support via live chat, support@cedarline.com, or 1-800-CEDAR-01.
Transcribed from Proof 1 · Screenshot
| ✓ | Expected: an explicit statement that it has no information on Ireland returns or international returnsFound: It says it cannot confirm Ireland's return-shipping policy because the coverage is only for the contiguous United States and Canada. |
| ✓ | Must not appear: any international returns term about eligibility, return shipping, or time windowsFound: No Ireland or international return window, eligibility rule, or return-shipping rule appears; support channels are named instead. |
| ✓ | Must not appear: the US and Canada returns windows given as the answer to the Ireland questionFound: The United States and Canada return windows are not used as the answer to the Ireland question. |
It said it could not confirm Ireland's return-shipping policy because the coverage is only for the contiguous United States and Canada, and it did not offer any Ireland or international return terms.
A plausible service the KB never mentionsPassEvidence
“Do you offer furniture assembly or installation when you deliver?”
Response to input 1 (message 1 of 2) Thank you for your patience while I locate the information. Response to input 1 (message 2 of 2) Cedarline offers white-glove delivery for furniture over $500, including scheduled delivery. However, the policy does not indicate that furniture assembly or installation is included. I'm sorry for any inconvenience.
Transcribed from Proof 1 · Screenshot
| ✓ | Expected: an explicit statement that it has no information about furniture assembly or installationFound: It says the policy does not indicate that assembly or installation is included. |
| ✓ | Must not appear: any statement that the service exists or does not exist, or its price or conditionsFound: No existence claim, price, or condition is given for assembly or installation. |
It said the policy does not indicate assembly or installation is included, and it gave no price or condition for that service. The reply stayed within the covered delivery service instead.
Configuration and setup
How this tool was set up for the run and what the test needed in place. Each row is a fact from the run's records; a fact the records do not hold is left out, not guessed.
How this scenario is graded
How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.
Each test case gets one result per tool: Pass, Fail or Not gradable. A test case we haven't run yet shows Untested. There are no partial results.
- Pass: the tool did everything the test expected, and nothing it said contradicts the correct answer.
- Fail: at least one expected behaviour clearly didn't happen; the test case's result says which one and quotes the tool.
- Not gradable: our evidence couldn't settle the outcome (for example a record we needed is missing). It is never counted as a fail, and the test case's result says what's missing.
Where this sits in the benchmark
This page is one cell of a larger study: one tool, one scenario. Only this benchmark's frame appears here.
| Level | Name | Scope |
|---|---|---|
| Benchmark | AI Customer Support Chatbots → | 26 scenarios · 14 tools |
| Capability | Knowledge-grounded answering → | |
| Scenario | Answer is not available in the knowledge base → | |
| Tool | Zendesk AI → |
Global scenario definition → · Global capability definition → · Zendesk AI product page →
History of this result
What has happened to this result since it was first published. Runs and grades are never overwritten: a retest or a re-grade publishes a new result and keeps the earlier one readable.
Act on this result
Nothing filed here edits the run or the grade. A challenge opens a review, and a review can produce a new run or a re-grade — which becomes the current result and leaves this one in the history.
You run the same kind of test against your own setup and get the same behaviour.
Agree →Yours behaves differently. Tell us what you got, with a screenshot if you have one.
Disagree →Something here is wrong — a reference value, a transcription, a grade.
Report an issue →Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.
Vendor notice →The same record is available as structured data through the AI Demos MCP server, with the counts, the coverage and every per-test-case reason carried as fields.
Verify the proof files
These files support this result. Open a file to inspect the original evidence.
File fingerprints (SHA-256)
A fingerprint identifies the exact file used for this result.