SHADOWFIREWALL
Architecte de défense adaptative
Profil professionnel
ShadowFirewall crée des pare-feu comportementaux fondés sur des modèles de comportements hostiles. Avec une expérience en IA et en systèmes embarqués, il conçoit des défenses réactives qui apprennent du trafic.
Compétences clés
Pare-feu adaptatifs; modélisation des comportements hostiles; détection d’anomalies; micro-segmentation; sécurité OT/SCADA
Réalisations majeures
Conception du premier pare-feu comportemental d’une usine intelligente européenne.
Articles de SHADOWFIREWALL
Identity Management Faces Its Next Test: When Non-Human Accounts Start Acting Like Operators
Non-human identities are already more numerous than human ones inside many enterprises, and the arrival of AI agents pushes identity control beyond static access lists toward governance of purpose, action, responsibility, and autonomy.
From Build Pipeline to SaaS Exposure: How a GitHub Actions Flaw Could Have Exposed Internal Credentials
A public workflow issue in Snowflake’s snowflake-connector-net repository illustrates how CI/CD script injection can turn routine automation into a secret-harvesting opportunity.
Encrypted Password Vaults Are Quietly Becoming Identity Defenses, Not Just Storage Apps
A discounted Proton Pass Plus offer puts the spotlight on a bigger security shift: password managers now bundle encryption, aliases, 2FA, dark web monitoring, and passkey support into one identity-control layer.
The Seven-Hour Secret Hunt That Exposed How Fast Cloud Defenses Can Collapse
A reconstructed attack path showed how one exposed credential, plus overbroad access and weak segmentation, can turn a single mistake into a wide blast radius across cloud and production systems.
The Quiet Identity Shift Behind Enterprise AI Risk
As AI agents move from chat interfaces into real workflows, security teams are being pushed to govern them like identities, not just software features.
Cloud Portal Misread as a Breach, Data-Extortion Claim Follows
A suspected Power Pages configuration issue is tied to a large data exposure claim, but the available evidence still points to public-table access rather than a confirmed exploit.
When the Directory Becomes the Target: Why AD Hardening Still Decides the Outcome
A new 2026 guide on Active Directory security points back to a familiar truth: if identity is weak, the rest of the environment inherits the risk.
Microsoft Pushes Entra ID Toward Passkeys, and Legacy OTPs Toward the Exit
Microsoft plans to make passkeys the default in Entra ID in 2026, then retire its own SMS and voice MFA delivery in 2027, forcing enterprises to confront the weakest link in their sign-in stack.
When the Inbox Is Not the Entry Point: Google Workspace and the OAuth Blind Spot
Google Workspace attacks may begin with stolen OAuth tokens, giving attackers another path into Gmail, Drive, and connected systems.
The Hidden Power Grid Behind Enterprise Logins
Identity and Access Management is often treated like admin plumbing, but in modern enterprises it is the system that decides who gets in, what they can touch, and how quickly access can be revoked.
Missing MFA Turns a Simple Credential Spray Into Ransomware Risk
A reported gap on remote access helped a credential-spraying attempt move into an Akira-linked intrusion, underscoring how one control failure can change the stakes fast.
Public Portals, Private Data: The New Harvesting Playbook on Salesforce and ServiceNow
A reported campaign called City-Forum is focusing on anonymous access paths in SaaS portals, where weak guest permissions can turn a public page into a quiet collection point.
When Identity Checks Move Into Chat, the Trust Problem Moves With Them
An in-WhatsApp verification experience can cut friction for users, but it also concentrates identity risk inside a channel attackers already love to imitate.
When a Low-Privilege Login Becomes Fleet-Wide Power
A critical Rancher authorization flaw puts the management plane itself in the crosshairs, where one authenticated account can turn into a much larger Kubernetes risk.
How a Brief LiteLLM Package Compromise Could Reach Deep Into Cloud Trust
A reported PyPI compromise involving LiteLLM shows how one poisoned release can become a supply-chain problem for CI/CD, Kubernetes, registries, and AI gateways.
Passkeys Under Pressure: Why a Phishing-Resistant Login Can Still Become a Target
A reported attack family aimed at Windows 11 and Microsoft Entra ID is a reminder that passwordless security depends on the whole identity chain, not just the cryptography inside the key.
Microsoft Cuts Off New Picture Password Setup in Windows 11
The change narrows one old sign-in path while reinforcing Microsoft's wider push toward PINs, passwords, and biometrics for Windows identity.
The Passkey Myth Crack: How Attackers Can Work Around Phishing-Resistant MFA
Fresh research shows that passkeys can remain cryptographically sound while the surrounding endpoint and sync layers still offer room for abuse.
Two Quiet Trust Layers, One Loud Warning: Identity Keys and NAT State Are Now in the Crosshairs
A pair of disclosures points at the same uncomfortable truth: attackers do not always need to break passwords or firewalls when they can target the long-lived state underneath them.
Phone Calls Are Becoming the Front Door to Cloud Account Theft
A tracked cybercrime crew tied to vishing is being associated with corporate account abuse, with Microsoft 365 and Okta in the spotlight and extortion risk following close behind.