Air Canada argues it is not liable for what its chatbot said; the tribunal rejects this, British Columbia, 14 February 2024

Changes to what this case says, with the reason. Every earlier version is kept.

• 2026-10-21: Inference: signed by. The interpretation's wording changed after the editor signed it (second, independent check of 21 october 2026); the signature is cleared so the editor signs the corrected wording.
  Show the change

  Before: Saurabh Bhatnagar

  After:
• 2026-10-21: Inference: value. Found in the editor's second, independent check of the interpretations, 21 october 2026: 'did not act on its own' could be read as 'was not autonomous', which contradicts the case's autonomy field; the decision shows the chatbot gave information and the customer made the bookings (2024 bccrt 149, paragraphs 14 and 18).
  Show the change

  Before: A boundary case: the chatbot did not act on its own. It is in the Register for the tribunal's answer to the airline's argument that it could not be held liable for what its chatbot said: the airline is responsible for all the information on its website, whether it comes from a static page or a chatbot. The ruling is a small-claims decision of British Columbia's Civil Resolution Tribunal. The record does not show how the chatbot worked or why it gave a wrong answer

  After: A boundary case: the chatbot only gave information, and the customer booked the flights. It is in the Register for the tribunal's answer to the airline's argument that it could not be held liable for what its chatbot said: the airline is responsible for all the information on its website, whether it comes from a static page or a chatbot. The ruling is a small-claims decision of British Columbia's Civil Resolution Tribunal. The record does not show how the chatbot worked or why it gave a wrong answer
• 2026-10-20: Applicable rules. Found in the editor's source re-check of 19-20 october 2026: the decision decides the claim in common-law negligent misrepresentation only (paras. 25-32, citing queen v. cognos inc.); the statutes it cites are the civil resolution tribunal act and the court order interest act. no consumer-protection statute is cited, argued or applied, and no source read supports one.
  Show the change

  Before: ["Negligent misrepresentation (common law)", "Consumer protection (provincial)"]

  After: ["Negligent misrepresentation (common law)"]
• 2026-10-20: Inference: value. Found in the editor's source re-check of 19-20 october 2026: the decision (para. 27) holds 'it should be obvious to air canada that it is responsible for all the information on its website' and 'it makes no difference whether the information comes from a static page or a chatbot'; its header gives the type as 'small claims'. no source read calls it a binding precedent or supports 'which every later agent case assumes', so both are removed; the decision gives no information about how the chatbot worked (para. 14), which the interpretation now says. the interpretation was not signed (signed_by null) before this change, so no signature is cleared; the new wording still needs the editor's sign-off.
  Show the change

  Before: A boundary case: the chatbot did not act on its own. It is in the Register for the precedent it set, that a company cannot disown what its automated system tells customers, which every later agent case assumes

  After: A boundary case: the chatbot did not act on its own. It is in the Register for the tribunal's answer to the airline's argument that it could not be held liable for what its chatbot said: the airline is responsible for all the information on its website, whether it comes from a static page or a chatbot. The ruling is a small-claims decision of British Columbia's Civil Resolution Tribunal. The record does not show how the chatbot worked or why it gave a wrong answer
• 2026-10-20: Controls: c-d1. Found in the editor's source re-check of 19-20 october 2026: the decision does not describe how air canada's chatbot worked or whether its answers were checked: para. 14 says 'air canada did not provide any information about the nature of its chatbot' and para. 28 finds 'air canada did not take reasonable care to ensure its chatbot was accurate'. the record supports 'no check is shown', not that no check existed, so confidence is lowered from high to medium; the control stays in the absent bucket (the editor's call).
  Show the change

  Before: {"note": "Answers that state policy are not checked against the airline's policy page before they reach the customer", "confidence": "high"}

  After: {"note": "The record shows no check of the chatbot's policy answers against the airline's policy page before they reached the customer; the tribunal found Air Canada did not take reasonable care to ensure its chatbot was accurate, and Air Canada gave no information about the nature of its chatbot", "confidence": "medium"}
• 2026-10-20: Fields: recovery technical. Found in the editor's source re-check of 19-20 october 2026: the field said 'none', but the decision (para. 18) records bookings on 11 and 16 november 2022 'relying on the information provided by the chatbot', and (para. 22) that an air canada representative 'admitted the chatbot had provided "misleading words"' and that 'air canada had noted the issue so it could update the chatbot'; business standard (19 february 2024) reported that 'air canada's chatbot support appeared to be disabled on its website' after the ruling. rule 16: a 'none' must cite a source that says so.
  Show the change

  Before: {"value": "None; the misstatement was not corrected before reliance", "source_ids": ["BCCRT"]}

  After: {"value": "None before reliance: the record shows no correction before the customer booked both flights relying on the chatbot's answer (11 and 16 November 2022). Afterwards, in February 2023, an Air Canada representative admitted the chatbot had provided \"misleading words\" and said Air Canada had noted the issue so it could update the chatbot; the decision does not say whether it was updated. After the ruling, Business Standard reported that Air Canada's chatbot support appeared to be disabled on its website.", "source_ids": ["BCCRT", "BUSINESS-STANDARD"]}
• 2026-10-20: Fields: outcome. Found in the editor's source re-check of 19-20 october 2026: the order in the decision (para. 44) is 'a total of $812.02', made up of $650.88 in damages, $36.14 in pre-judgment interest and $125 in tribunal fees; the field said 'damages and fees of roughly cad 812' and left out the interest.
  Show the change

  Before: State: harm (small). The customer was refused the retroactive discount; tribunal awarded damages and fees of roughly CAD 812. As of 2024

  After: State: harm (small). The customer was refused the retroactive discount; the tribunal ordered Air Canada to pay CAD 812.02 (CAD 650.88 damages, CAD 36.14 pre-judgment interest, CAD 125 tribunal fees). As of 2024
• 2026-10-20: Fields: decision. Found in the editor's source re-check of 19-20 october 2026: the chatbot's words as reproduced in the decision (para. 15) were 'if you need to travel immediately or have already travelled' and 'within 90 days of the date your ticket was issued'; the 90 days ran from ticket issue, which the field now says.
  Show the change

  Before: The chatbot stated that a bereavement fare could be claimed retroactively within 90 days, contradicting the airline's own policy page

  After: The chatbot stated that a reduced bereavement fare could be claimed after travel, within 90 days of the date the ticket was issued, contradicting the airline's own policy page
• 2026-10-20: Context: autonomy level text. Found in the editor's source re-check of 19-20 october 2026: the decision (moffatt v. air canada, 2024 bccrt 149, para. 14) says 'the parties implicitly agree that mr. moffatt was not chatting with an air canada employee' and 'air canada did not provide any information about the nature of its chatbot'; it says nothing about human review, so 'no human review' is restated as what the record shows rather than as a finding.
  Show the change

  Before: Machine answers alone; no human review of individual answers

  After: Machine answers alone (the parties agreed the customer was not chatting with an Air Canada employee); the record shows no human review of individual answers
• 2026-10-14: Controls. Verifier rule 10: both candidates describe the same mechanism as c-d1 (pre-execution validation); a control is recorded in one state per case.
  Show the change

  Before: candidates 'Consistency between chatbot and policy source' (failed) and 'Grounding of chatbot answers in the authoritative policy' (absent); C-D1 note 'Human review of policy-bearing answers'

  After: folded into C-D1 (absent) with a clarified note
• 2026-10-13: Fields. Independent check against the linked records before right-of-reply letters.
  Show the change

  Before: title and claim said Air Canada argued its chatbot was a separate legal entity

  After: Air Canada's argument in its own terms; the 'separate legal entity' characterisation attributed to the tribunal; Air Canada's point about the linked policy page added