CASE / 0317
UNRESOLVED
Cross-border consumer accountability
Bridging the Redress Gap.
AI-assisted accountability for vulnerable cross-border consumers. AI organizes the evidence. People make the call.
Click, swipe or use ←→
Evidence gapMissing opening video
LiabilityWho is responsible?
Next actionHuman review required
334,500cross-border requests · 2022
+42.63%growth since 2020
8 + 2cases + expert interviews
01
A dispute should not end
The redress gap
A dispute should not end
before it begins.
Low-value cross-border claims can cost more to resolve than the product itself. Responsibility, evidence and route become three separate barriers.
02
The problem is
The evidence
The problem is
measurable.
The concept is grounded in consumer complaints, cross-border trade data, public cases and expert interviews.
12315 CROSS-BORDER REQUESTS
2022
334,500
Cross-border-related requests recorded in a single year.
YEAR-ON-YEAR GROWTH
vs 2020
42.63%
1.47× the 2020 level.
CROSS-BORDER SEIZURE SHARE
2025
63.6%
24,600 of 38,700 infringement-related seizures.
RESEARCH EVIDENCE BASE
PROJECT
8cases
Plus two expert interviews totalling 2h 29m.
30 legal & policy sources
7 design questions
2 expert interviews
1 evidence-led concept
03
Follow the
The method
Follow the
evidence trail.
Five research actions turned legal rules, case patterns and expert insight into one testable accountability workflow.
Design synthesis
Rules → evidence → workflow
04
AI is useful.
What we found
AI is useful.
It is not a judge.
Select a glass card to reveal the evidence behind each finding.
05
From uncertainty to an
The solution
From uncertainty to an
actionable complaint.
A guided path that keeps decisions human.
HUMAN INPUT
Step 01 / 04
Describe the dispute
Consumers add the order, chat, purchase-time page snapshot, logistics and product media through a low-friction interface.
Plain-language prompts
Consent before collection
Pause and resume anytime
AI-assisted
Human judgment remains the final checkpoint.
Appeal-ready
06
Transparency is
Trust by design
Transparency is
part of the product.
The concept should explain what is missing, not pretend to know what is true.
HUMAN
IN THE LOOP Consent · Review · Appeal
IN THE LOOP Consent · Review · Appeal
No unsupported claims. Rule-engine limits constrain AI discretion at every step.
07
The tools are ready.
Conclusion
The tools are ready.
The judgment stays ours.
This project maps a lower-barrier, evidence-led route to redress while keeping authentication, liability and legal judgment outside AI.
AI remains outside final judgment.
Thresholds, data access and partner uptake need testing.
3–5 users · 8-case back-test · partner interviews · prototype.
AI organizes the evidence.
People make the call.