Algeria's customs authority launched ALCES, a clearance platform whose risk module uses AI to flag import/export fraud risk. It processed 120,000 declarations and registered 6,000+ operators in nine months, later expanding to travelers — but no fraud-catch rate is published.
Direction Générale des Douanes (Algerian Customs), Ministry of Finance
Period
November 2023–present
Keywords
customs, trade facilitation, fraud detection, digital government
Context
Algeria's Direction Generale des Douanes (the customs directorate, under the Ministry of Finance) launched ALCES in November 2023 to replace paper-based customs declarations. One module applies artificial-intelligence-based techniques to analyze import/export risk indicators and flag shipments for fraud and tax-evasion checks. In June 2025, the platform was extended to travelers, who can now declare currency and valuables online before reaching the airport or port.
Results
By July 2024, the Ministry of Finance reported that ALCES had processed more than 120,000 customs declarations and registered over 6,000 economic operators across Algeria's land, sea, and air customs offices. Independent outlet Maghreb Emergent covered the rollout and confirmed the risk-management module's AI framing, but neither the Ministry of Finance nor independent press have published a fraud-detection rate, processing-time comparison, or independent audit of the AI component specifically — the public record so far describes function and volume, not measured impact.
Implementation
Indicative cost
Medium (€50k–€500k) — No budget figure is published; classified as medium cost for a nationwide customs-declaration platform with an AI risk-scoring module.
Time to results
Medium (1–3 years) — Launched November 2023 and extended to traveler declarations in June 2025, about a year and a half of rollout.
Staffing & skills
Direction Generale des Douanes (Algerian Customs), Ministry of Finance
Conditions for success
Digital platform replacing paper-based declarations across land, sea and air customs offices
Risk module scores pre-submitted declaration data
Phased expansion from declarations to traveler currency/valuables declarations
Common failure modes
No fraud-detection rate, processing-time comparison, or independent audit of the AI risk module has been published
Risk-scoring criteria and any appeal mechanism for flagged declarations are not disclosed
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.
Azerbaijan Customs built ARAS, a machine-learning risk system sorting shipments into green/yellow/blue/red corridors. The customs chief reported a 58% cut …