Video summary

Comment retrouver n'importe qui sur Internet ? (Sensibilisation OSINT)

Main summary

Key takeaways

Educational

Main ideas, concepts, and lessons

  • Online anonymity is an illusion: The video argues that being “anonymous behind your screen” doesn’t hold up once identifiers are reused and data from different sources is linked together.
  • OSINT (Open Source Intelligence): Introduces OSINT as a discipline used by investigative journalists, intelligence services, and cybercriminals—focused on searching and correlating public information rather than hacking.
  • No malicious hacking (but serious tracing risk): The walkthrough is presented as legal OSINT intended to demonstrate how easily personal data can be assembled and how to protect oneself.
  • User/data reuse creates an investigation path:
    • Using the same nickname/username everywhere (“user reuse”) is described as the core vulnerability.
    • Combining username → accounts → location hints → email → map/geolocation → identity is portrayed as a step-by-step “loop closure.”
  • Geolocation via imagery and metadata: A mundane social post (photo) plus shadows, architecture, and mapping tools can narrow from city-level to a specific building.
  • Email is “ultimate digital identity”: Reverse-searching an email can reveal connected Google services, reviews, and location-relevant patterns.
  • French administrative transparency: In France, the video claims it can be easier to connect people to real identities via indexed public company/association records.
  • Phishing threat scenario: If an attacker knows accurate personal details, they can craft highly convincing targeted phishing (e.g., “hello Guillaume, your package at …”).
  • Protective guidance (“digital hygiene”):
    • Compartmentalization (separate addresses for leisure vs. serious use).
    • Don’t reuse usernames across platforms.
    • Regular self-checking (search your own data using OSINT tools).

Practical methodology / step-by-step workflow (as demonstrated)

0) Setup: use safer searching conditions (anti-leak framing)

  • Rule: Don’t do OSINT searches directly from your personal computer or home network.
  • Reason: Clicking malicious links or exposing your IP could make the investigator identifiable (“hunter becomes prey”).
  • Demonstrated approach: Use a VPS (described as a disposable/isolated “digital bunker”) so requests come from the VPS IP.

1) Start with a username (initial clue)

  • Input example: A fictional nickname found in a YouTube comment (e.g., “xx retrogamer”).
  • Goal: Expand the username into other online handles to discover more linked accounts.

2) Find other accounts using username reuse

  • Step A: Run “Sherlock”
    • A script/tool that queries many websites for the same username.
    • Output is used to identify sites where the username appears.
    • Note: may yield false positives and may miss sites if blocked.
  • Step B: Check “What’s My Name”
    • A web database to search the username and find associated profiles.

3) Pivot: use the most informative discovered platform(s)

  • Concept: “pivot”
    • Move from information set A to set B that opens new leads (e.g., from social profiles to marketplace location).
  • Example pivot in the video: Vinted
    • Marketplace profiles can include a location.
    • Extract location like city (Lyon) from the profile.
    • This narrows search scope from the whole world to a single city.

4) Extract contact identifiers (email) from forums/social bios

  • Step: Check an older forum account profile for a listed email address.
  • Claim: The video finds an email and treats it as a key identity anchor.

5) Reverse-search the email to reveal linked services and location signals

  • Tool introduced: Epios (email reverse lookup described as “vicious/brilliant”).
  • Process:
    • Enter the email into Epios.
    • It queries Google-linked services to find associations.
  • Specific outputs highlighted:
    • Public Google Calendar (if exposed)
    • Google Maps reviews left by the user
  • Inference: “pattern of life”
    • If reviews exist for multiple nearby businesses (e.g., within ~300m), the person may live or work near the center of that triangle.

6) Automate/extend discovery with additional scripting (OL)

  • Tool introduced: “OL” (name partially garbled in subtitles).
  • What it does (as described):
    • Automates checking “forgot password” on many popular sites using the target email.
    • Interpretation:
      • If it returns “unknown email” → likely not registered
      • If it sends a recovery email → likely registered
  • Outcome claimed: psychological/profile inferences
    • Example inference targets:
      • Strava → potentially athletic
      • Dating site → possibly single
      • Betting site → possibly gambler
  • Result: Add findings back into the Maltego graph for visualization.

7) Obtain more personal identifiers (phone number) via listings

  • Method described: location + vehicle/objects + local classifieds
    • If an old profile indicates a car model and location, search local classifieds (e.g., LeBoncoin).
    • Look for ads posted by someone with similar username/first name.
    • Contact/responses can reveal a phone number.
  • Also mentioned: The same logic applies to apartment rentals.

8) Geoint (geolocation): from city/building to exact building

  • Goal: Move from “Lyon (city/district)” to the exact building.
  • Step A: Use a Twitter photo
    • A photo with a view and shadows provides geometry cues.
    • Look for identifiable elements:
      • Street/route segment
      • Opposite-side buildings/architecture
      • A sign suggesting a tram/bus stop
  • Step B: Use Google Earth (3D mode)
    • Validate the building layout by matching the view.

9) Use shadows + time to identify the correct facade/floor context

  • Tool introduced: “SunCalc” (called “suncalk” in subtitles).
  • Method:
    • Use the photo’s date/time (from tweet or metadata).
    • Simulate the sun position and shadow direction.
    • Align with observed shadows to confirm the photo angle and point to the building facade.
    • Draw inference line from a shadow to the light source direction.

10) Close the loop: connect pseudonym to real identity via public records

  • Claimed France-specific advantage: Administrative/public corporate data is indexed.
  • Tools/websites mentioned:
    • Papers.com (subtitle mentions “papers.com”)
    • societe.com
  • Method (as described):
    • Use search queries on these services, sometimes through “Google dorks” style queries to find indexed documents.
    • Search by elements like name fragments, Lyon, street name, and/or email/username.
    • Download free documents (articles of association, general meeting minutes).
  • Final linkage example described:
    • Official document containing: full name, birth date, birth city, and current address.

11) Threat model: why attackers care

If malicious actors have real name, exact address, contextual details, and timestamps/behavioral patterns, targeted scams become more convincing (e.g., package delivery phishing).

Protection: “golden rules” recommended in the video (digital hygiene)

  • Compartmentalization
    • Use separate “junk/leisure” vs “serious” addresses for different purposes.
    • Keep banking/tax/official matters on a dedicated serious address.
  • Avoid username reuse
    • Don’t use the same nickname/handle everywhere.
  • Cleaning / self-audits
    • Regularly “google yourself” by searching your:
      • username
      • email
    • Run your email through reverse/OSINT tools (video specifically recommends Epios / similar tools referenced).
  • Optional community call
    • Join an “associationfr” community (mentioned as an investigative resource).

Speakers or sources featured (as named in subtitles)

  • Host / narrator: “Franzo” (speaker and channel/person presenting the video)
  • Sponsor mentioned: Hostinger (VPS provider)

Tools (named sources used in the demonstration)

  • Maltego
  • Sherlock
  • What’s My Name
  • Epios
  • OL (script for automated password-recovery/registration checking)
  • Overpass Turbo
  • Google Earth (3D mode)
  • Google Maps
  • SunCalc
  • Papers.com
  • Societe.com

Communities/services referenced

  • YouTube (video where the username was supposedly found)
  • Instagram
  • TikTok
  • Spotify
  • Tinder
  • Twitter
  • GitHub
  • Vinted
  • gbatemp.net
  • LeBoncoin
  • Google services (Calendar, Maps)

Original video