BrandVakt — BrandVakt: an autonomous agent that closes the 21-hour window phishing kits live in
Client · Cybersecurity
BrandVakt
ClientBrandVakt
SectorCybersecurity
IndustrySecurity & Trust
RegionEurope

BrandVakt: an autonomous agent that closes the 21-hour window phishing kits live in

Industry research shows 91% of breaches start with phishing and the median phishing kit stays live for 21 hours. BrandVakt is an AI agent built by HomoDeus to collapse that window: continuous scanning of the open web, LLM-based threat classification, automatic evidence packaging, and takedown initiation, all without waiting for a human triage queue.

Across 11 industry verticals served by BrandVakt, the operational reality is the same: 2,237 new phishing domains are registered every day on average (APWG), 7% of phishing emails slip past traditional filters (industry baseline), and security teams cannot scan the open web continuously. HomoDeus built BrandVakt as a productised AI agent that does exactly that, on behalf of enterprise brands, around the clock.

Outcomes

21 hrsmedian phishing kit lifespan before takedown (APWG industry baseline)
91%of breaches start with phishing (Deloitte industry research)
2,237new phishing domains registered per day on average (APWG)
7%of phishing emails slip past traditional filters (industry baseline)

Inside that 21-hour window, BrandVakt watches, classifies, and acts. Security teams stop being the bottleneck; humans approve exceptions and own strategic response.

The problem

Phishing kits go live in hours and die in hours; APWG benchmarks place the median lifespan at 21 hours. Reactive monitoring loses the window. Most enterprises buy a dashboard and still depend on a human analyst to triage and act, which means the analyst becomes the bottleneck. Meanwhile 91% of cyberattacks start with phishing (Deloitte) and traditional secure email gateways still let 7% of phishing through.

What we built

HomoDeus built BrandVakt as an autonomous agent platform: continuous scanning across newly registered domains, threat feeds, search engines, and dark web channels; LLM-based classification of phishing kits, brand abuse, and vulnerability disclosure patterns; automatic evidence packaging; takedown initiation against registrars and platforms; and a security team interface for review and approval where escalation is needed. Productised to deploy per enterprise brand without standing up a new SOC.

Stack

  • Continuous scanning engine
  • LLM threat classifier
  • Evidence packaging pipeline
  • Automated takedown integrations
  • Analyst escalation interface

Time to production

Productised agent, deployable per brand

Questions a C-suite leader will ask

What does BrandVakt do?

BrandVakt is an autonomous AI agent built by HomoDeus that continuously scans the open web for phishing kits, brand impersonation, lookalike domains, and exposed vulnerabilities. It classifies threats, packages evidence, and initiates takedowns without waiting for human triage.

How is BrandVakt different from a brand monitoring dashboard?

Dashboards wait for a human to look. BrandVakt is an agent that watches continuously, classifies threats with LLMs, packages evidence, and initiates response automatically. Security teams approve exceptions instead of triaging every alert.

Related case studies

           .