NEURALSHIELD
Ingénieur de protection des systèmes d’IA
Profil professionnel
Défend les systèmes d’IA contre des attaques telles que le poisoning et l’inversion.
Compétences clés
Sécurité de l’IA; mitigation du data-poisoning; défense contre l’inversion de modèle; analyse de jeux de données; durcissement des pipelines IA
Réalisations majeures
Arrêt d’une attaque qui aurait manipulé des millions de décisions automatisées.
Articles de NEURALSHIELD
Search Ads, Fake Installers, and the Mac Trust Trap
A malvertising campaign is using Google Ads and a fake Claude Code installer to reach macOS users, showing how cybercrime now leans on trust instead of obvious exploits.
The Quiet Trade-Off Behind “Spam Call” Cleanup
A discount on annual privacy-removal plans draws attention to a bigger question: how much can broker opt-outs really shrink the data trail behind unwanted calls and social engineering?
Support Calls, Remote Access, and a Hidden Backdoor: A Quiet Path Into Employee PCs
A reported social-engineering intrusion used fake IT contact, Microsoft Teams, and Quick Assist to reach employee computers before the GoGRPC backdoor was installed, with ransomware ties still unconfirmed.
Fake Download Traps Turn Windows Search Into a Malware Magnet
A large impersonation campaign used clone domains, copied branding, and AI-written pages to make bogus Windows app sites look routine and trustworthy.
Insurance Phishing Has Learned to Move at the Speed of a Login
CTM360-linked research points to a shift from delayed credential theft to real-time account hijacking, a change that shrinks the defender’s window from hours to seconds.
Trusted Labels, Hidden Payloads: How a Fake Logistics Email Became a Malware Delivery Trap
A phishing run dressed up as shipping notices and tax-audit mail shows how attackers can turn everyday business trust into a credential-theft pipeline.
When a Login Code Becomes the Key: Snapchat Phishing Case Exposes the Weak Spot in MFA
A guilty plea tied to Snapchat code theft shows how one-time codes can fail when attackers aim at the user, not the platform.
When a Photo Becomes a Weapon: The Hidden Pipeline Behind Nudify Abuse
AI image-manipulation tools can turn ordinary photos into synthetic sexualized images, making consent the first casualty and pushing schools, families, and regulators toward infrastructure-level defense.
The Help Desk Trap: Fake IC3 Sites and AI Voices Are Turning Recovery Into a Second Scam
A new FBI warning shows how criminals are blending impersonation, synthetic media, and lookalike complaint portals to catch fraud victims at their most vulnerable.
The Fake Microsoft Page That Chases the Session, Not Just the Password
AiTM phishing turns a trusted login screen into a relay point for credentials, MFA output, and live session tokens, which can make account takeover look like a normal sign-in.
The Cheapest Crime in Cybersecurity Still Works: Why Phishing Keeps Slipping Past the Guardrails
Phishing stays effective because it abuses trust at scale, moving through email and other channels while defenders are forced to combine awareness, filtering, and stronger authentication.
FaceTime Scams Turn Live Video Into a New Trust Trap
Fraudsters are using FaceTime calls to impersonate banks, tech support, and government officials, making old social engineering tactics feel more immediate and harder to question.
Three Evilginx Campaigns Put Microsoft 365 Identity Security Under a Sharper Light
A reported exposure of attacker infrastructure adds a rare defensive clue to three phishing campaigns built around Microsoft 365 login abuse.
One Loose Server, Three Phishing Cells: How a Public Python Port Unraveled an Evilginx Network
A small operational mistake on a public host exposed live Microsoft 365 phishing infrastructure, revealing how fragile attacker tradecraft can be when shell history and directory listings are left in plain sight.
The Microsoft Login Trick That Turns a Real Sign-In Into a Phishing Win
A long-running device-code campaign shows how attackers can abuse a legitimate Microsoft authentication flow to harvest access without breaking in the usual way.
Brazil’s Security Budgets Rose. Phishing Found the Weakest Door Anyway.
A long-running survey of Brazilian companies shows stronger security investment, yet phishing still climbed sharply, underscoring how identity abuse can keep working even when defenses mature.
Forg365 Turns Microsoft 365 Phishing Into a Bought-and-Sold Access Business
A Telegram-linked phishing service shows how identity theft now borrows the mechanics of SaaS, combining device-code abuse, token persistence, and AI-written lures.
When Antivirus Starts Hunting Lies: Avast’s AI Pitch Shows Where Scam Defense Is Headed
A discount may be the headline hook, but the technical story is the shift from malware blocking to detecting phishing, fake websites, and synthetic media before users act.
The Phone Call Was the Trap: Robinhood-Style Alerts Turn Trust into a Weapon
A callback-phishing campaign uses fake sign-in warnings to push recipients away from inbox checks and into attacker-controlled voice channels.
Phone Calls, Fake Managers, and the Cloud Trail to SharePoint
A reported extortion crew called Helix illustrates how voice phishing, device-code abuse, and MFA weak spots can turn identity trust into a path toward SharePoint data.