Korpalis

AI security intelligence you can actually verify.

Korpalis cuts through AI-generated noise. A curated research database, hands-on consulting and a weekly newsletter, where every single claim links back to a real paper, CVE or official source.

Create free account Preview the research
12 verified sources monitored
8 AI threat categories
100% of claims linked to a source
Illustration of an AI neural network protected by a security shield, with threat vectors blocked at the edge
Sources we monitor arXiv NVD / CVE OWASP GenAI MITRE ATLAS CISA SANS PortSwigger NIST
What we do

Three ways in, one standard of proof

The newsletter keeps you current. The database is where you go with questions. Consulting is for answers too specific to find in an article.

01

Research Database

A curated, searchable library of AI security research. Hook summary first, depth on demand.

  • 8 threat categories, from prompt injection to red teaming
  • 3 difficulty levels: Beginner, Intermediate, Advanced
  • Every entry backed by a paper, CVE or official doc
Preview the database →
02

1-on-1 Consulting

Booked calls of 60 to 90 minutes, live or async code review, and written follow-up notes.

  • Security-lens reviews of AI and vibe-coded apps
  • Agent architecture and production hardening
  • Tool selection: gateways, guardrails, RAG security
Book a session →
03

Weekly Newsletter

A free, human-written digest of AI security news, research and governance updates.

  • Incidents and research explained in plain language
  • EU AI Act and NIST updates without the legalese
  • No sponsors, no filler, no vendor pitches
Get the newsletter →
How we work

Built on source integrity

The EU AI Act and NIST AI RMF are pushing companies toward AI governance whether they are ready or not. Defensible references matter more than ever.

Every claim has a source

If we cannot link it to a real paper, CVE or official document, we do not publish it. Simple as that.

Written for your level

Beginner gets the takeaway, Intermediate gets the mechanism, Advanced gets the full paper. Never patronising, never overwhelming.

Always current

AI security moves weekly. We track 12 verified sources so you do not have to trawl arXiv on your Sunday evenings.

Who it is for

From solo builders to boardrooms

Solo developersShipping with AI without shipping its mistakes
Software engineersStaying current without reading arXiv weekly
CISOs and consultantsCitable references for clients and boards
Product teamsDeploying AI with strategy and controls
Free account

Start reading today

Create a free account for the current research, sorted by topic and difficulty. Upgrade when you want the full archive, search and alerts. No card needed, no spam, ever.

Create free account Or just the weekly newsletter