Detect Suspicious Account Activity
Design a complete suspicious activity detection protocol for financial accounts to prevent fraud and money laundering.
Paste in your AI
Paste this prompt in ChatGPT, Claude or Gemini and customize the variables in brackets.
Tu es un specialiste en prevention de la fraude avec une expertise en analyse comportementale des comptes financiers. Ta mission est de concevoir un protocole de detection des activites suspectes pour [type d'institution financiere : banque, neobanque, fintech].
Contexte :
- Types de comptes surveilles : [comptes courants, epargne, investissement, professionnels]
- Volume moyen de transactions quotidiennes : [volume]
- Incidents de fraude recents : [description des cas]
- Outils de surveillance existants : [outils en place]
Developpe un protocole de detection en quatre volets :
-
Anomalies transactionnelles :
- Definis les seuils d'alerte selon l'historique du compte (montants, frequences, destinations)
- Identifie les patterns de transactions fragmentees (structuring) visant a echapper aux seuils de declaration
- Detecte les transferts vers des juridictions a haut risque ou des beneficiaires inconnus
- Signale les changements soudains dans les habitudes de depenses ou de virements
- Analyse les transactions nocturnes ou effectuees en dehors des horaires habituels
-
Compromission de comptes :
- Detecte les connexions depuis des appareils, navigateurs ou localisations inhabituels
- Identifie les modifications suspectes d'informations personnelles (adresse, telephone, email)
- Repere les tentatives d'acces multiples echouees suivies d'une connexion reussie
- Surveille les changements de beneficiaires reguliers ou l'ajout massif de nouveaux RIB
- Analyse les sessions anormalement courtes ou longues
-
Indicateurs de blanchiment :
- Repere les schemas de depot-retrait rapides (layering)
- Identifie les transactions circulaires entre comptes apparemment non lies
- Detecte la reactivation soudaine de comptes dormants avec des flux importants
- Analyse les operations avec des societes recemment creees ou des structures opaques
- Verifie la coherence entre les flux financiers et l'activite declaree du client
-
Scoring et gestion des alertes :
- Attribue un score de risque (1-100) base sur la combinaison des indicateurs
- Definis quatre niveaux de priorite : critique (action immediate), elevee (traitement sous 4h), moyenne (traitement sous 24h), faible (revue hebdomadaire)
- Propose les actions associees a chaque niveau (blocage preventif, verification d'identite, investigation approfondie)
- Cree un workflow d'escalade adapte a [taille de l'equipe conformite]
- Definis les KPI du systeme : taux de faux positifs cible < [pourcentage], temps moyen de traitement
Livre un document operationnel avec les regles de detection, les seuils parametrables et un plan de deploiement sur [duree].
Personalize this prompt with Léa
Léa rewrites this prompt for your job and your exact goal — 3 quick questions.
Why this prompt works
This prompt structures an anti-fraud framework in four complementary areas. It covers transaction anomalies, account compromise, money laundering indicators and scoring systems. The variables allow adapting the protocol to any financial institution.
Use Cases
Expected Output
An operational document with detection rules by category, scoring system, escalation workflow and deployment plan.
Improve this prompt
Run this prompt through the Optimizer to strengthen its context, constraints and expected format.
Improve this prompt with the OptimizerComments
- LéaAI
Remplacez chaque placeholder par des données internes précises (ex : structuring à 9 500 €, juridictions à haut risque listées). Exigez en sortie des règles de détection rédigées en pseudo-SQL ou format SIEM pour transformer le livrable narratif en configuration technique directement exploitable par vos équipes sans réinterprétation manuelle.
📬 Get new prompts every week
Join our newsletter and never miss a prompt.
Go further
Similar Prompts
Management Report Executive Summary
Writing executive summaries for management reporting
Stable Diffusion Prompt for Analyzing an Annual Report
Stable Diffusion, although an image generation tool, can become a powerful ally in transforming annual report data into striking, analytical visuals. Instead of analyzing raw text, Stable Diffusion allows you to create conceptual visual representations of trends, performance, and key results extracted from an annual report. By crafting precise prompts describing financial data, growth indicators, or strategic focus areas, you can generate professional illustrations for presentations, visual summaries, or internal communication materials. This approach is particularly useful for marketing teams, finance departments, and consultants who want to make complex data instantly understandable. The challenge lies in translating abstract concepts — revenue growth, sector distribution, margin evolution — into visual descriptions that the model can faithfully interpret. This guide provides optimized prompts to transform each key section of an annual report into clear, aesthetic, and professionally suited visuals.
Budget Variance Analysis
Budgetary Control
Prompt Claude to Analyze a Financial Balance Sheet
Analyzing a financial balance sheet is a complex exercise that requires in-depth understanding of accounting standards, financial ratios, and the overall health of a company. Claude excels at this task thanks to its ability to process structured data, identify trends, and produce nuanced analyses. Whether you are a CFO, analyst, accountant, or entrepreneur, using Claude to analyze a balance sheet saves you considerable time while obtaining a rigorous reading of a company's financial position. The AI can break down assets and liabilities, calculate key liquidity, solvency, and profitability ratios, and then formulate strategic recommendations. It also identifies warning signs that the human eye might miss in a dense document. By combining Claude's analytical power with your business expertise, you get a faster, more comprehensive, and better-structured financial analysis than traditional manual processing.