SHADOWFIREWALL
Adaptive Defense Architect
Professional Profile
ShadowFirewall creates behavioral firewalls based on hostile-pattern models. Background in AI and embedded systems, designs reactive defenses that learn from traffic.
Key Skills
Adaptive firewalls; Hostile-behavior modeling; Anomaly detection; Micro-segmentation; OT/SCADA security
Major Achievements
Designed the first behavioral firewall of a European smart factory.
Articles by SHADOWFIREWALL
When an AI Agent Logs In, Who Owns the Access?
Enterprise AI is turning software into a new class of identity problem: a delegated actor that can move, ask, and act inside corporate systems, but still needs tight governance, traceable authority, and fast revocation.
When the Login Is Strong but the Recovery Path Is Not
A new identity lifecycle service is aimed at the weakest stretch of modern access control: proving who someone is before they get a credential, and again when they need it back.
When Automation Identities Leak, MSP Trust Can Turn Into a Blast Radius
A reported campaign tied to stolen application credentials in managed service provider tenants shows how cloud automation can become a cross-tenant risk, even when the full impact is still unconfirmed.
Entra’s Hidden Risk Surface: Why Identity Add-Ons Matter More Than Their Marketing
A comparison of Microsoft Entra security add-ons points to the same pressure points defenders keep rediscovering: legacy authentication, privileged access, identity proofing, hardware-backed credentials, and recovery after something goes wrong.
When a Workload Identity Turns Into a Back Door
A reported Entra ID credential campaign shows how one compromised application identity can ripple across tenants, mailboxes, and downstream customers without a human login ever being obvious.
Hijacked AI Access Can Burn Through Cloud Budgets Before Anyone Notices
Reported 2026 incidents show how unauthorized use of enterprise LLM access can turn identity compromise into a billing emergency, with charges reportedly topping $100,000 in a day.
Federal Access Partnerships Look Strong on Paper - The Hard Part Starts After the Announcement
An expanded partnership may support more controlled federal access, but the path from product announcement to resilience is not automatic.
When AI Access Becomes a Meter Running Wild
Credential leaks can enable LLMjacking, where attackers abuse enterprise access to AI models and drive up cloud bills.
When Digital Certificates Stop Living on Shared Rails
Mind Group reportedly moved certificate handling from a shared remote server to centralized control, a change that can tighten governance without automatically proving stronger security.
SASE Gets Another Spotlight as Cloud Security Teams Revisit the Edge
Hackers Online Club published an explainer on Secure Access Service Edge, a reminder that the architecture is as much about policy design as it is about moving controls into the cloud.
Sovereign Cloud Is More Than Geography: The Hidden Battle Over Access, Law, and Control
A cloud can sit in Europe and still raise sovereignty questions if access, jurisdiction, and governance are unclear.
Palo Alto’s Unified SaaS Push Bets on a Single View of Fractured Risk
Palo Alto Networks has launched a platform aimed at bringing SaaS, cloud, identity and AI risk into one operational picture, a move that may help analysts spot connections they would otherwise miss.
When Cloud Settings Become the Crime Scene: CISA’s Baseline Tools Expose the Weakest Link in SaaS Security
Two assessment utilities for Microsoft 365 and Google Workspace turn tenant settings into policy checks, but they also show why configuration hygiene is only one layer of defense.
Passkeys Close One Door, Not the Whole House
Passwordless login can blunt phishing, but the real security test starts around enrollment, recovery, sessions, and the device itself.
NIST Draws a Hard Line Around AI Agents: Identity, Privilege, and Audit Trails
NIST has proposed a blueprint for treating AI agents as accountable identities, not just software tools - a move that could reshape how enterprises control permissions, delegation, and review.
When a Critical Bug Meets a Thin Permission Set, the Damage Changes Shape
PHP2Shell is being used as a warning about speed: patching matters, but limiting what service accounts, API keys, and AI agents can reach may matter just as much when defenders are racing the clock.
Microsoft’s Passkey Deadline Signals the End of an Old Identity Shortcut
Entra ID administrators are being pushed toward phishing-resistant sign-in methods as Microsoft prepares to retire its own SMS and voice authentication path on a staggered timetable.
The Quiet Leak Is Often the Permission You Forgot You Gave
A buyer’s guide on data access governance lands at the center of a bigger problem: AI assistants can turn ordinary over-permissioning into instantly searchable exposure.
When Cloud Settings Become the Weakest Link, CISA’s Free Scanners Step In
ScubaGear and ScubaGoggles do not block attacks on their own, but they can expose the misconfigurations that make Microsoft 365 and Google Workspace easier to abuse.
Approval Is Not Access: The Cloud Log Allegation That Exposed a Bigger Trust Problem
Screenshots linked to a disputed access workflow may hint at a review step, but they do not by themselves prove that government cloud logs were actually reached or misused.