DIVISION 02 · CYBERSECURITY · OFFENSIVE SECURITY

Blackglass. Break it before someone else does.

An AI-assisted web application pentesting platform that combines deterministic testing, strict policy controls and evidence-based verification — safer automation, higher-quality findings, none of the scanner noise.

01WHAT IT DOES
Machine-scale coverage
AI drives the breadth, determinism drives the trust. Test plans expand with the attack surface — but every probe is reproducible and every run replayable.
Policy-fenced
Strict scope and policy controls decide what the platform may touch and how hard — automation that cannot wander outside the engagement.
Evidence, not guesses
Findings ship with verification evidence — request, response, proof of impact. If it can't be demonstrated, it doesn't get reported.
SaaS or self-hosted
Multi-tenant SaaS, or self-hosted in your own Docker/Kubernetes environment when the targets can't leave the network.
02HOW IT WORKS
01 · Scope
Targets, rules of engagement and hard limits are encoded as policy before the first packet — the fence is code, not a PDF.
02 · Test
Deterministic modules plus AI-guided exploration cover the application — every step logged, every mutation replayable.
03 · Verify
Candidate findings are re-executed and verified against evidence before they reach the report — the false-positive pile stays with the machines.
04 · Report
Findings ranked by demonstrated impact, with reproduction steps your engineers can run — and re-run after the fix.

When did you last test like an attacker?

Engagements by agreement, for vetted organisations.

Get in touch