French is spoken by over 300 million people across five continents—in France, Quebec, Belgium, Switzerland, and throughout Africa. Each French-speaking region has distinct identifier systems, address formats, and linguistic characteristics. Effective French data protection requires understanding these variations: the NIR in France, NAS in Quebec, numéro national in Belgium, AVS in Switzerland. Our French language processing provides native understanding across the entire Francophone world.
Beyond identifier formats, French presents unique linguistic characteristics for PII detection. Naming conventions include particles (de, du, de la), compound first names, and distinct patterns by region. Address formats differ between European and North American French contexts. Our French NLP models are trained specifically on French data, understanding these nuances rather than relying on translation from English patterns.
French Social Security Number (NIR)
The Numéro d'Inscription au Répertoire, commonly called the numéro de sécurité sociale, is France's primary personal identifier:
Structure: The NIR is a 15-digit number (13 digits plus 2-digit control key) encoding personal information:
- Digit 1: Gender (1 = male, 2 = female)
- Digits 2-3: Year of birth (2 digits)
- Digits 4-5: Month of birth (01-12)
- Digits 6-7: Department of birth (or 99 for overseas)
- Digits 8-10: Commune of birth (INSEE code)
- Digits 11-13: Registration number
- Digits 14-15: Control key (calculated from digits 1-13)
Validation: The control key is calculated as 97 minus (the 13-digit number modulo 97). We validate this control key to confirm valid NIRs while rejecting random 15-digit sequences.
CNIL Requirements: The CNIL has specific rules limiting NIR use to certain purposes (social security, healthcare, employment). Detection helps identify potentially improper NIR collection or retention.
French Naming Conventions
French names have distinctive characteristics requiring specialized detection:
Name Particles: Many French surnames include particles: "de" (Jean de Lattre), "du" (Pierre du Pont), "de la" (Marie de la Fontaine), "des" (Jacques des Essarts). These particles are part of the surname and should be detected together.
Compound First Names: French commonly uses compound first names: Jean-Pierre, Marie-Claire, Jean-Baptiste, Anne-Sophie. These function as single given names, not first and middle names.
Hyphenated Surnames: Modern French usage includes hyphenated surnames from marriage: Martin-Dupont, Bernard-Lambert. Detection handles these as complete surnames.
Regional Variations: Names reflect regional heritage—Breton names (Yann, Gwenaël), Alsatian names (Hans, Liesel), Corsican names (Ange, Toussaint). Our NLP recognizes these regional patterns.
French Address Formats
French addresses follow specific conventions:
Standard Format: French addresses place the street number before the street name: "25 rue de la Paix." Street types include rue, avenue, boulevard, place, impasse, chemin, and many others.
Code Postal: The 5-digit postal code precedes the city name. First two digits indicate the department (75 = Paris, 69 = Rhône, 13 = Bouches-du-Rhône). This enables department identification from postal codes.
CEDEX: Business addresses often include CEDEX (Courrier d'Entreprise à Distribution EXceptionnelle), a special postal service. CEDEX addresses have specific formats with CEDEX numbers.
Building Identifiers: French addresses may include building designations (Bâtiment A), staircase (Escalier 2), and apartment numbers (Appartement 45).
Quebec French
Quebec French has distinct characteristics from European French:
Vocabulary Differences: Quebec uses different terms: courriel (email) vs. courrier électronique, stationnement (parking), magasinage (shopping), and numerous Anglicisms unique to Quebec.
Quebec Identifiers: The Numéro d'assurance sociale (NAS/SIN) is Canada's social insurance number. Quebec's RAMQ (Régie de l'assurance maladie du Québec) issues health insurance numbers with a specific format. Driver's licenses follow Quebec-specific patterns.
Address Formats: Quebec addresses use North American conventions: street number before name, but with French street types (rue, avenue, boulevard). Province abbreviation is QC in English contexts, Québec in French.
Bilingual Context: Quebec documents often include both French and English elements. Our processing handles code-switching within documents.
Belgian French
Belgium's French-speaking region has specific identifier systems:
Numéro National: Belgium's national registry number (also called numéro de registre national or NISS for social security) is an 11-digit identifier encoding birth date and gender with a control number. Format: YY.MM.DD-XXX.CC
Belgian Addresses: Belgian addresses use postal codes (4 digits) with municipalities. The address format resembles French conventions but with Belgian-specific elements.
Swiss French
Switzerland's French-speaking cantons (Romandie) have Swiss-specific identifiers:
AVS/AHV Number: The Swiss social insurance number (numéro AVS in French, AHV in German) is a 13-digit identifier with the format 756.XXXX.XXXX.XX. The 756 prefix indicates Switzerland.
Swiss Addresses: Swiss addresses use 4-digit postal codes (NPA - numéro postal d'acheminement). Canton abbreviations (GE, VD, VS, NE, JU, FR) identify regions.
Financial Identifiers
French financial data has specific formats:
French IBAN: French IBANs use format FR + 2 check digits + 23 characters (5-digit bank code + 5-digit branch code + 11-character account + 2-digit RIB key). Total length: 27 characters.
SIREN/SIRET: French business identifiers. SIREN is 9 digits identifying a company; SIRET adds 5 digits (NIC) for establishment location, totaling 14 digits.
French Phone Numbers: French numbers use 10 digits typically displayed as 0X XX XX XX XX. Geographic codes: 01-05 for landlines by region, 06/07 for mobile. International format: +33 X XX XX XX XX.
GDPR/CNIL Compliance
France implements GDPR through CNIL enforcement:
CNIL Guidance: The CNIL publishes detailed guidance on personal data categories, security measures, and processing requirements. Our detection aligns with CNIL definitions of personal and sensitive data.
NIR Restrictions: CNIL specifically restricts NIR use to authorized purposes. Detection helps ensure NIR isn't collected or retained beyond authorized uses.
Data Minimization: GDPR's minimization principle, emphasized by CNIL, encourages limiting personal data collection. Redaction operationalizes this by removing unnecessary personal data.
African Francophone Countries
French is official in many African nations, each with specific identifier systems:
Regional Variations: Morocco, Algeria, Tunisia, Senegal, Côte d'Ivoire, and other Francophone African countries have national ID systems. Our detection covers major African Francophone identifiers.
Address Formats: African addresses vary significantly by country. Major cities have street addressing; rural areas may use descriptive locations. Processing adapts to available address information.
