An AI capability, used in-house
Cortex is the AI capability that sits behind our team. It helps us read faster, draft faster, sift more signal from more noise, and free up the part of every engagement that has to be human. We use it on our own work, on customer work, and on the things we publish.
When customer data is involved, Cortex runs on hardware we own and configurations we control — not somebody else's API. That's a structural decision, not a marketing promise. The full version of how that works is further down.
It is not a product. We don't license it, white-label it, or sell access to it. The reason for that is on this page too — it's worth saying out loud.
Where Cortex shows up
Concrete uses, not vibes. Every place Cortex appears, a human still owns the outcome.
-
01
Research & threat analysis
Reading the long tail of advisories, blog posts, and reports so analysts can spend their time on the parts that matter for a specific customer environment.
-
02
Grid alert triage assistance
First-pass enrichment and de-duplication on incoming alerts — never the final word, always handed to a human for the call that gets escalated.
-
03
Education content & translation
Helping move our compliance courses across 29 languages without losing meaning, and drafting new modules for human review and edit.
-
04
Consulting deliverables
First drafts of assessments, policies, and reports — structured by us, drafted with assistance, then rewritten and signed off by the consultant who owns the work.
-
05
Internal tooling
Glue code, log parsers, one-off scripts. The unglamorous engineering that keeps a small team punching above its weight.
The honest version
The security industry is mid-rush to package AI as a product. Most of what's getting sold is somebody else's model, wrapped in a UI, marked up several times, and pointed at a problem the buyer hasn't fully scoped. We've watched it. We're not interested in adding to it.
Cortex earns its keep by making our team better at the work we already do. That value goes to customers in the form of faster, sharper, cheaper engagements — not as a line item on an invoice. If AI capability is what you actually need, there are real options out there. We'll happily point you at them.
“A tool we use, not a product we ship.”
Guardrails, in plain language
If you're a customer, this is what you can expect. If you're not, this is what we hold ourselves to anyway.
On hardware we own
When customer data is involved, Cortex runs on our own machines, in our own racks, on configurations we control — not on a third-party API. Your data doesn't leave the perimeter we run, doesn't train anyone's model, and isn't retained beyond the task. The structural answer to "where did my data go" is: nowhere it shouldn't.
-
01
No autonomous action on your environment
Cortex doesn't take actions on a customer's systems on its own. Anything that touches your environment goes through a human.
-
02
Disclosure on request
We'll tell you, on any engagement, where Cortex helped and how. No mystery boxes. If you'd rather we not use it on your work, that's fine too — just say so.
-
03
The human is on the hook
Every output that leaves our team has a person's name on it. The consultant, the analyst, the engineer — they own the call, not the model.
Questions about how we use AI?
Ask. We'll answer plainly — what tools, what data, what we won't do, and where the line between assistance and decision sits.
> info@censoredsystems.com