THE SECURITY BRUTALIST

Security Unconventional Warfare Part 1: Cutting through the noise

Most breaches trace back to neglected fundamentals, not brilliant adversaries. Organizations chase shiny solutions while missing obvious gaps in their own foundations, and three habits explain most of that pattern. Stacking overlapping tools on top of each other creates integration problems, since each new platform needs its own tuning, its own maintenance, and its own analyst training, and none of that upkeep work makes an attacker's job any harder. Vanity metrics point leadership toward numbers that look good in a slide deck, like phishing emails blocked, while a stale privileged account or an unsegmented network keeps sitting untouched in the background. And a security team drowning in fifty thousand daily alerts has no realistic way to triage that many, so a domain admin compromise can pass by unremarked in the middle of routine notifications. Attackers depend on exactly this kind of disorder; a system left off the inventory, or a data store without proper controls, is often all they need to get through.

Security Brutalism answers this with fewer priorities and fiercer focus. Rather than build broader coverage, it narrows a program down to four laws that build on each other in sequence, and each law rests on a pair of practical disciplines underneath it.

Know what you have. A defender cannot protect an asset they do not know exists, and an attacker only needs the one system nobody remembers, so asset management starts with a living inventory of hardware, software, and accounts, reconciled on a real schedule rather than dusted off once a year before an audit. Data security sits right alongside it, since information is the actual target of nearly every attack, and that means encryption at rest and in transit, sensitivity classification, and access mapped to who actually needs it, with intellectual property and customer records treated as the crown jewels they are.

Make it hard to break. Identity and access do this work first, removing stale credentials, enforcing least privilege without carve-outs for seniority, and requiring verification at every boundary, executives included, since even the strongest segmentation becomes useless once a high-privilege account bypasses it. Segmentation extends the same idea to the network itself, building zones of differing value with logged choke points between them, so a foothold in one zone does not become a highway to every other.

See trouble fast. No defense is perfect, so speed of detection becomes the next best thing, and that runs through visibility built on high-signal logging, favoring depth over breadth and focused on authentication, privilege escalation, and boundary crossings rather than collecting everything possible.

Limit and recover. When something does get through, patch and vulnerability management is what keeps the damage contained, closing known weaknesses on a schedule tracked to an accountable owner, internet-facing systems first, so every incident handled well leaves the organization more resilient than it found it.

These four laws reinforce each other in a loop, since each one depends on the law before it to actually work. Inventory gives hardening something real to harden, a hardened baseline gives detection something to measure against, detection gives containment a signal to act on, and containment gives the whole program something to recover from cleanly. Skip one law and the weakness does not disappear, it just surfaces later as a failure in whichever law gets tested next.

My old red team ran into exactly this during an assessment of a large enterprise with a mature security department, the kind of organization that looked well-defended on paper. The team reached domain admin within a few hours, and the post-mortem showed exactly which law had gone unenforced first, since the asset inventory listed a fraction of the real device count, customer data sat exposed on forgotten file shares behind default passwords, and segmentation existed only in architecture diagrams rather than in the actual network. Once inventory was wrong, hardening had nothing solid to protect, and once hardening was soft, detection and containment never stood a chance. Leadership cut a stack of redundant tools in the first month and rebuilt the inventory from scratch, and alert volume dropped from tens of thousands a day down to a couple hundred, with every one of those worth acting on. Six months later, the same red team came back and got caught rather quickly. The fix cost nothing in new budget. It came from doing fewer things and doing them without exception.

None of this holds without a culture that treats the fundamentals as non-negotiable. Technology fails without discipline, because attackers do not need creativity when defenders are inconsistent, and one poorly configured admin account can invalidate thousands of monitoring rules just as easily as one unpatched system left "for convenience" can undermine millions of dollars of investment. The industry has built sophisticated detection algorithms and threat intelligence feeds for years, and most breaches still succeed through basic failures.

Discipline breaks down in predictable ways. Executives ask for exceptions around MFA or segmentation, claiming their role requires it, and operators comply out of fear rather than judgment, which is exactly why executive accounts become high-value targets. Under pressure, staff disable controls "for speed" and promise to restore them later, but the pressure never really ends and the shortcut becomes permanent. Policies exist on paper while enforcement stays weak, so audits check documentation instead of practice and create false confidence, and teams end up chasing reporting formats and vanity metrics instead of closing the gaps that would actually stop an attacker.

Building a disciplined culture takes coordinated effort at three levels. Leaders need to say plainly that fundamentals come before any new initiative, repeat that message often, and refuse privilege exceptions for themselves as much as for anyone else, since a bypassed MFA requirement at the top signals that rules are optional for the powerful. Quarterly reviews of actual practice, rather than annual paperwork checks, keep pressure on standards, and recognition should go to the steady work of clean asset inventories and consistent patch cycles as much as to dramatic incident response.

Teams need to fold fundamentals into everyday routines rather than treating them as separate projects, running identity checks as part of onboarding, building vulnerability scanning into deployment pipelines, and running asset discovery continuously instead of once a quarter. Exceptions need to surface immediately rather than get hidden until someone notices, and peer review on critical changes keeps shortcuts from becoming permanent under pressure. Individuals need to treat repetitive tasks such as rechecking logs and reviewing access as central professional work rather than administrative overhead, take ownership of gaps without waiting for an external audit, and speak up when someone else bypasses policy, since silence lets the culture erode for everyone.

Enforcement needs teeth. Metrics on closure rates and review completion should be visible across the organization, since transparency creates its own peer pressure, and internal audits should test actual configurations and access logs rather than reviewing documents alone. Persistent failure should carry consequences, including role reassignment when necessary, while a genuine mistake gets treated differently from repeated disregard, and training works best through drills that show how one unpatched server can open the door to a full network compromise, building the muscle memory that holds up under real stress.

Culture drifts without maintenance. Quarterly reinforcement sessions, rotating staff through fundamental responsibilities, and consistent leadership messaging keep the standard from eroding as new projects compete for attention. Discipline is what turns policy into practice, building a kind of reliability that resists exploitation even as attacker tools grow more sophisticated.



Go to Part 2.