AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
Graded 17 October 2026

Does Zendesk AI follow configured instructions that change its behavior?

Zendesk AI followed the configured instruction in this return chat. The instruction was active, and the customer’s candle-return request got a polite refusal that named Home Decor and did not start or approve a return.

1 of 1 test case passed

Every test case in this scenario has a result.

Pass rate100%1 of 1 gradable
Coverage1 of 1test case with a result
1PassDid everything it was expected to do.
0FailNo test case failed.
0Not gradableEvery result here could be graded.
0UntestedEvery test case in this scenario has a result.
The pass rate is a summary. The evidence is the test case below: what we sent, what we checked, what the tool returned, and the proof.

The test case

Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on this test case.

Configured behavioral instruction appliedPassEvidence
What we sent

“I already opened the candle I ordered — can I still return it?”

What the tool returned

Unfortunately, candles are Home Decor items and are not eligible for return, even if unopened. I apologize for any inconvenience or confusion this may cause.

Transcribed from Proof 1 · Screen recording at 0:28

Every message the tool sent, in order (2)
  1. Progress message

    Thank you for your patience while I find the answer.

  2. Answer

    The final response, printed above.

Expected vs. Found
✓Expected: The answer follows the configured instruction.Found: The agent treated the candle as Home Decor, refused the return politely, and started no return.
Supporting proof
Proof 1Screen recording
Screen recording: the active instruction and the return chat reply
Key moments
Open original ↗
Why this result

It is a Pass because the customer asked to return a candle, and the agent treated candles as Home Decor, refused the return politely, and started no return.

Also observed in this test case

Holding message. The reply is preceded by a brief progress message.

Sources chip. A Sources chip appears under the answer.

Tested by Ajay Vekhande · evidence dated 29 September 2026

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.

Software that produced the output
Zendesk AI
Build or version
Not exposed by the hosted product
Plan or tier
Suite - Professional Free trial
Surface
Web app
Set up before the run
A customer request about an existing order or item was available, and one active instruction was configured to shape the reply.
Instruction
Home decor items are non-returnable
Instruction status
One active; last updated 29 September 2026 at 5:21 PM
Tested
29 September 2026 · Ajay Vekhande

How this scenario is graded

How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.

How results are decided

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.

The rules
  • 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.

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.

from the publication record
13 November 2026First publishedAI Customer Support Chatbots v1

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.

This matches what I see

You run the same kind of test against your own setup and get the same behaviour.

Agree →
This does not match

Yours behaves differently. Tell us what you got, with a screenshot if you have one.

Disagree →
Point out an issue

Something here is wrong — a reference value, a transcription, a grade.

Report an issue →
Request a retest

On a newer build, a larger dataset, or your own setup.

Request a retest →
We have fixed this

Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.

Vendor notice →
Filed against this evidenceNothing yet. Challenges, counter-evidence and fix notices appear here with their outcome, and stay on the page after they are resolved.
Cite this result
aidemos.com/benchmarks/ai-customer-support-chatbots/results/zendesk/configured-instruction-changes-agent-behavior · 1 pass · 0 fail · coverage 1/1 · graded 2026-10-17

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.

Proof 1 · Screen recording6ee60c3702a17f0fa9d28c79635b57aef7bf3ff52f345f785b8ae66a1d77cd98