evidoria

← Back to browse

Good practice Imported

Malaysia's Customs Department Pairs AI Scanners With a Record RM1.88 Billion in Seizures

Malaysia · Putrajaya · See the Malaysia profile · See the Putrajaya profile

Evidence: Observational / pre–post Top 83% 40/100 · Ask Evidence Copilot about this practice

JKDM deployed 134 AI-equipped scanning machines at border checkpoints in 2025, the year its seizures of smuggled and duty-evaded goods rose 15.3% to a record RM1.88 billion across 7,097 cases — though the agency's own accuracy claims are self-reported.

134
AI-equipped scanning machines deployed
1.88 billion RM
2025 seizures value (2025)
7,097
2025 seizure cases (2025)
1.63 billion RM
2024 seizures value (2024)
15.33 %
Year-on-year seizure increase
52.92 million RM
KLIA AI-chip seizure (June 2026)
65
Behavioural-detection officers trained
Malaysia's Customs Department Pairs AI Scanners With a Record RM1.88 Billion in Seizures

Details

Maturity
Scaling
Promoter
Jabatan Kastam Diraja Malaysia (Royal Malaysian Customs Department, JKDM)
Period
2025
Keywords
customs, border security, smuggling detection, computer-vision scanning

Context

The Royal Malaysian Customs Department (JKDM) has deployed AI-based image-recognition scanning machines at border checkpoints, paired with a 'Neuro-Linguistic Programming' behavioural-detection initiative to flag suspicious travellers.

Objectives

Increase detection of smuggled and duty-evaded goods at borders and entry points.

Activities

134 scanning machines identify scanned goods within 20–30 seconds, while 65 trained officers apply behavioural-detection techniques alongside the scanners.

Results

2025 seizures of smuggled and duty-evaded goods reached RM1.88 billion across 7,097 cases, up 15.33% from RM1.63 billion across 7,215 cases in 2024; the largest categories were cigarettes (RM497.27 million) and drugs (RM490.75 million), and a separate June 2026 sweep at Kuala Lumpur International Airport intercepted a RM52.92 million shipment of misdeclared AI server chips.

Conclusions

These figures, and JKDM's claim of '100% accuracy' for the scanners, are self-reported by the department through state media and have not been independently audited; the seizure increase cannot be attributed to the AI scanners alone since it coincides with wider enforcement operations.

Implementation
Indicative cost
High (€500k–€5M)
Time to results
Medium (1–3 years)
Staffing & skills
65 officers trained in Neuro-Linguistic Programming behavioural detection, JKDM checkpoint scanning operators

Conditions for success

  • Combining AI scanners with trained human behavioural detection
  • Deployment across border checkpoints and entry points nationwide

Common failure modes

  • JKDM's '100% accuracy' claim is a self-reported departmental assertion, not an independently verified metric
  • Seizure increases cannot be isolated from wider, concurrent enforcement operations

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful