Tool in benchmark · Version 1

Zendesk AI in AI Customer Support Chatbots

Scenario-level performance from current published Results.

18Passed
4Failed
3Different test set
1No published Result yet

25 of 26 scenarios published

CapabilityScenariosResult
Action execution4 pass · 0 fail · 5 of 5 published
Agent configuration and control2 pass · 0 fail · 2 of 2 published
Analytics and observability3 pass · 0 fail · 3 of 3 published
Conversation continuity2 pass · 1 fail · 3 of 3 published
Escalation and human handoff3 pass · 0 fail · 3 of 3 published
Feedback and learning0 pass · 2 fail · 2 of 2 published
Intent understanding3 pass · 0 fail · 3 of 3 published
Knowledge-grounded answering0 pass · 0 fail · 2 of 3 published
Multilingual support1 pass · 1 fail · 2 of 2 published
PassedFailedPublished, different test setNo published Result yet

Where Zendesk AI failed

4 of 22 graded scenarios. Open a Result to see the conversation and the evidence.

Conversation continuity

Earlier conversation/session needs to be continued

Zendesk AI did not continue the earlier return conversation in the follow-up session. In the second session, the agent answered generically and then gave support channels instead of naming the earlier order or sofa, while the earlier exchange stayed in the same thread.

View Result →
Feedback and learning

Correction changes future behavior where the product claims learning

Zendesk AI could not carry out the correction-and-re-ask task because the tested agent had no way to accept a correction. The question was never asked again in a new conversation.

View Result →
Feedback and learning

Feedback can be submitted

Zendesk AI offered no way to submit a correction to one of its earlier answers on the conversation screen where those answers are read. That screen's only menu held an export option, the conversation overview offered only a field for adding labels, and no rating, flag or correction option appeared on the agent's replies.

View Result →
Multilingual support

Customer switches languages during the conversation

Zendesk AI failed the language-switch scenario. After the customer switched to English, it kept the policy context but answered in French, including the follow-up reply and the holding line before it.

View Result →

How Zendesk AI performed

Open a capability to explore its scenarios. Each row reports the test set in its published Result.

Action execution4 pass1 different test set
ScenarioPublished outcomesTest coverageResult
Action requires confirmation
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI asked for confirmation before cancelling the subscription. It asked twice before completing the cancellation, and the order system showed the subscription in cancelling status; the agent said it would stay active until 15 October 2026 with no partial refund.

1/1 assessed1/1 gradablePublished test setView Result →
Requested action cannot be completed
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI refused the address change request on the shipped order and did not claim success. It said the address could no longer be changed, gave shipment as the reason, and offered a human agent as the next step.

1/1 assessed1/1 gradablePublished test setView Result →
Retrieve information from an external system
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI retrieved information from an external system and reported the returned order status. For order 41982, the conversation log shows the Check Order Status action firing successfully and returning shipped, UPS, tracking number 1Z999AA10123456784, and estimated delivery 2026-09-22; the customer reply repeats those same fields. The recording shows the action firing and returned fields, but not the HTTP request line or endpoint.

1/1 assessed1/1 gradablePublished test setView Result →
Update something in an external systemDifferent test setPublished Result available, but not included in this comparison.View Result →
User is not authorized to perform the action
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI refused a lookup for an order that belonged to another customer and did not reveal its details. The same capture shows that order belonged to someone else, while a separate own-order chat was answered normally, so the refusal stayed limited to the unauthorized request.

1/1 assessed1/1 gradablePublished test setView Result →
Agent configuration and control2 pass
ScenarioPublished outcomesTest coverageResult
Configured instruction changes agent behavior
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

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.

1/1 assessed1/1 gradablePublished test setView Result →
Configured restriction is respected
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI declined the delivery-address change without offering to make it. It explained that delivery addresses on existing orders cannot be changed and offered to check the order status using an order number.

1/1 assessed1/1 gradablePublished test setView Result →
Analytics and observability3 pass
ScenarioPublished outcomesTest coverageResult
Agent actions/tool calls can be inspected
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI passes this observability check: it shows the order-lookup call and lets the operator inspect it. The conversation lists triggered actions as successful, and opening the action details shows the action name, time, and success snapshot; the conversation is also addressable by URL.

1/1 assessed1/1 gradablePublished test setView Result →
Conversation activity can be inspected
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI listed an earlier shipped-order conversation in its own conversation list and opened it with the full transcript readable: the customer's order question, the order-status action marked Success, and the agent's shipping reply. Its action details panel also showed the order data that action used.

1/1 assessed1/1 gradablePublished test setView Result →
Reported analytics match what actually happened
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI's reporting overview for 23 September to 15 October 2026 showed 50 conversations and 11 assisted escalations, and all 25 of our earlier test conversations, including the three escalation tests, appear in its conversation logs. Its totals contain our runs; the extra conversations are other traffic on the same agent.

1/1 assessed1/1 gradablePublished test setView Result →
Conversation continuity2 pass1 fail
ScenarioPublished outcomesTest coverageResult
Customer corrects information given earlier
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI answered about the corrected order 41830 and did not carry over 41982. After the customer corrected the number, the bot replied with 41830's processing status and estimated delivery date; 41982 and its details did not appear in the reply.

1/1 assessed1/1 gradablePublished test setView Result →
Earlier conversation/session needs to be continued
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Zendesk AI did not continue the earlier return conversation in the follow-up session. In the second session, the agent answered generically and then gave support channels instead of naming the earlier order or sofa, while the earlier exchange stayed in the same thread.

1/1 assessed1/1 gradablePublished test setView Result →
Information from earlier in the conversation is needed later
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI used an earlier order number to answer a later delivery question without asking for it again. It correctly identified UPS as the carrier and reported that the order was shipped and not yet marked delivered.

1/1 assessed1/1 gradablePublished test setView Result →
Escalation and human handoff3 pass
ScenarioPublished outcomesTest coverageResult
A configured rule requires escalation
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI passed this escalation scenario. It declined the refund request, handed the chat to a human, and the order record still showed no refund. The capture also shows the ticket open with a human reply.

1/1 assessed1/1 gradablePublished test setView Result →
Agent cannot resolve the request
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI passed this scenario: it said it could not answer the account-change request and routed it to a human. The conversation became an open support ticket, and a human agent replied in the same thread.

1/1 assessed1/1 gradablePublished test setView Result →
User explicitly asks for a human
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI handled the request to speak to a human and passed it to a live agent. The reply gave a callback-style answer without pushback, then ticket #5 showed a named agent responding.

1/1 assessed1/1 gradablePublished test setView Result →
Feedback and learning2 fail
ScenarioPublished outcomesTest coverageResult
Correction changes future behavior where the product claims learning
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Zendesk AI could not carry out the correction-and-re-ask task because the tested agent had no way to accept a correction. The question was never asked again in a new conversation.

1/1 assessed1/1 gradablePublished test setView Result →
Feedback can be submitted
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Zendesk AI offered no way to submit a correction to one of its earlier answers on the conversation screen where those answers are read. That screen's only menu held an export option, the conversation overview offered only a field for adding labels, and no rating, flag or correction option appeared on the agent's replies.

1/1 assessed1/1 gradablePublished test setView Result →
Intent understanding3 pass
ScenarioPublished outcomesTest coverageResult
One message contains multiple requests
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI answered both requests in one reply to a message that asked for order status and return policy. It said order 41982 had shipped with UPS, and it gave the 21-day standard return window, the 45-day Cedarline Plus window, and the resalable-condition/original-packaging guidance.

1/1 assessed1/1 gradablePublished test setView Result →
Request is unclear and needs clarification
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI asks for clarification when a request is unclear. In the case shown here, it asked which cancellable thing the customer meant and did not cancel anything.

1/1 assessed1/1 gradablePublished test setView Result →
Request requires choosing the correct source or action
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI passed this order-status case. The conversation shows one custom action between the customer’s question and the reply, and the reply matches the order system’s shipped status, UPS carrier, tracking number, and estimated delivery date.

1/1 assessed1/1 gradablePublished test setView Result →
Knowledge-grounded answering2 different test set1 no Result yet
ScenarioPublished outcomesTest coverageResult
Answer is available in the knowledge baseDifferent test setPublished Result available, but not included in this comparison.View Result →
Answer is not available in the knowledge baseDifferent test setPublished Result available, but not included in this comparison.View Result →
The request uses different wording than the sourceNo published result
Multilingual support1 pass1 fail
ScenarioPublished outcomesTest coverageResult
Customer communicates in another supported language
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Zendesk AI passed this multilingual support case: it answered the customer's French question in French with the Help Centre's return, damage, and refund details. Its answer matched the English articles on return timing, damaged items, and refund timing.

1/1 assessed1/1 gradablePublished test setView Result →
Customer switches languages during the conversation
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Zendesk AI failed the language-switch scenario. After the customer switched to English, it kept the policy context but answered in French, including the follow-up reply and the holding line before it.

1/1 assessed1/1 gradablePublished test setView Result →

Reading these Results

Published evidence and test coverage answer different questions.

Publication availability

Which scenarios have a Result?

A published Result is public evidence for this tool on one scenario. “No published result” does not say whether testing has taken place.

Test coverage

What does each Result cover?

Assessed includes Pass, Fail and Not gradable. Gradable includes Pass and Fail. Both use the pinned test count in that published Result.

Scenario scope

Out of 26 scenarios

The summary counts each of the 26 scenarios once. A scenario fails when any of its published test cases failed. “No published Result yet” means nothing is public for that scenario; it does not say whether Zendesk AI was tested on it.

Zendesk AI in AI Customer Support Chatbots | AI Demos