Obstruction · Australia

Dead end

Dead end is a working label for this design mechanism: A consumer-favourable route cannot reach its represented outcome and instead loops, stalls or loses progress. This is editorial implementation guidance for Australian journeys, not a finding of unlawfulness. From 1 July 2027, a similar interface is relevant to ACL section 28B only if the complete consumer-connection, manipulation or unreasonable-distortion and actual-or-likely-detriment test is met. Current ACL rules require a separate assessment.

Editorial implementation guidance
Also known as
  • interaction dead end
  • blocked path
  • decline route returns to same prompt
  • no completion control
  • error only on consumer-favourable path
  • back navigation loses progress

Definition

What is this pattern?

A consumer-favourable route cannot reach its represented outcome and instead loops, stalls or loses progress. This is a design and research taxonomy for learning and evidence review, not a statutory offence label or an automatic finding that an interface is unlawful.

How it works

A consumer-favourable route cannot reach its represented outcome and instead loops, stalls or loses progress. A route represents that a request can complete but then loops, errors or changes no durable state, without a working recovery path or confirmation record.

Warning signs

  • The interface represents or reasonably implies that the outcome is available.
  • Repeated execution cannot reach that outcome under controlled conditions.
  • A business-favoured route remains functional or the failure systematically burdens the alternative.

Potential harms

  • A customer may abandon a valid request after repeating work without a way to complete or recover.
  • The represented outcome is not reached, leaving the subscriber exposed to another charge.

Learn by comparison

What does this look like?

These fictional examples make the design mechanism easier to recognise. They do not depict a real company and do not establish that an individual interface is unlawful.

Illustrative example 1 · Refund form loops after submission

A fictional refund form accepts every field, then returns to the first screen with “Something went wrong” and no saved state or alternative route.

Potential consumer harm: A customer may abandon a valid request after repeating work without a way to complete or recover.

Illustrative example 2 · Cancellation confirmation changes no account state

A fictional service shows “Cancellation complete” but returns the user to an active-plan dashboard and continues renewal without a receipt or status event.

Potential consumer harm: The represented outcome is not reached, leaving the subscriber exposed to another charge.

What is a fairer alternative?

Provide a functional completion route, preserve progress, and surface recoverable errors equally across options.

Context matters

Context and boundary cases

  • The interface represents or reasonably implies that the outcome is available.
  • Repeated execution cannot reach that outcome under controlled conditions.
  • A business-favoured route remains functional or the failure systematically burdens the alternative.
  • Boundary to test: General outage affecting all paths
  • Boundary to test: validation error caused by invalid fixture input
  • Boundary to test: clearly unavailable option

When a similar design can serve a legitimate purpose

  • General outage affecting all paths
  • validation error caused by invalid fixture input
  • clearly unavailable option

Operational review

What teams should review

Teams
  • Product
  • Design
  • Engineering
  • Legal
  • Compliance
  1. How many steps, waits and channel changes separate “Try again” from the completed order support outcome?
  2. Which fields or permissions are required, what happens on refusal, and does the resulting state support this criterion: “The interface represents or reasonably implies that the outcome is available”?
  3. Measure the same task through the clearest available route: does the effort difference persist once “General outage affecting all paths” is accounted for?
  4. How many steps, waits and channel changes separate “Back to account” from the completed subscription account outcome?
  5. Count steps, waits, offers and channel changes through final confirmation; where does the route meet “Repeated execution cannot reach that outcome under controlled conditions”?
  6. Measure the same task through the clearest available route: does the effort difference persist once “validation error caused by invalid fixture input” is accounted for?

Evidence to retain

  • Annotated account management and cancellation screenshots at each responsive breakpoint
  • The complete state sequence before, during and after the consumer decision
  • Design-system component, content, default and configuration records for the reviewed release
  • Operational records substantiating price, availability, timing and eligibility claims
  • Usability, accessibility, reversal, complaint and support evidence relevant to consumer impact
  • A dated product and legal review record identifying evidence, uncertainties and release decisions

Legal map and implementation tools

Evidence base

Sources

  1. An Ontology of Dark Patterns KnowledgeGray et al.; ACM CHI 2024 · Secondary · checked 2026-09-14 · DOI 10.1145/3613904.3642436; arXiv:2309.09640
  2. Competition and Consumer Amendment (Unfair Trading Practices) Act 2026Federal Register of Legislation · Primary · checked 2026-09-14 · C2026A00064
  3. Competition and Consumer Act 2010, including Schedule 2: Australian Consumer LawFederal Register of Legislation · Primary · checked 2026-09-14 · C2004A00109
  4. Unfair trading tricks and traps to be bannedTreasury Ministers · Primary · checked 2026-09-14
  5. Inquiry into the Competition and Consumer Amendment (Unfair Trading Practices) Bill 2026Senate Economics Legislation Committee · Primary · checked 2026-08-09