genai
SECURITY LAB
For consultancies & PTaaS

Clients are asking for AI pentests. Staff them.

The scope requests arrived before the skills did. Train your existing testers on live AI targets, certify the ones you put in front of clients, and show which of your bench can take the work.

Team is $39/seat/mo, billed yearly, with a 4-seat minimum. GSCE sittings are bought separately at $399 (or GSCP at $199). Launch prices until 27 November, then $499 and $249.

Individual tester rather than a firm?

Bench readiness

Know who you can put on the engagement

Coverage per tester across the OWASP LLM Top 10, MCP (Model Context Protocol) tool servers and agentic categories, with GSCE status alongside it. When a client asks who is doing the work, the answer is a record rather than an assurance.

genaisecuritylab.com/insights
Who can take an AI engagement
Bars: coverage across the OWASP LLM Top 10, MCP and agentic categories · sample data, not a real bench
MRMarcus R.GSCE
PNPriya N.GSCE
LOLena O.Not yet
TBTom B.Not yet
Assignments

Set a date and see who is ready by it

Assign a learning path or single labs to the testers who need them, with a due date. Completion is graded automatically, so no senior has to review each attempt, and the workspace shows who finished in time.

genaisecuritylab.com/team
Assignments
Learning paths assigned with due dates · sample data, not a real bench
Offensive AI Security3 of 4 done
Due in 5 days · Whole bench
MCP Boundary Testing1 of 2 done
Due in 12 days · 2 people
Hardening RAG0 of 1 done
Due in 19 days · Tom B.
Delivery

One methodology across every tester

New consultants follow the same path as your seniors, so review time does not scale with headcount.

genaisecuritylab.com/methodology
AI pentest methodology
1
Scope
Map the AI surface: models, tools, retrieval, agents
2
Map
Trace where untrusted text reaches context
3
Test
Work the OWASP LLM Top 10 against each entry point
4
Verify
Re-run the payload against the proposed fix
5
Report
Evidence, impact, remediation the client can action
Win the scope

Turn down fewer AI engagements. Your testers already know web and cloud — this closes the gap that loses the bid.

Certify the bench

GSCE — the GenAI Security Certified Expert — is graded on live targets, proctored through the candidate's webcam, and reviewed by an examiner before it is issued. Each certificate has a public verification URL a client can check, so it is worth putting on a consultant's profile in a proposal.

A method, not improvisation

A repeatable AI pentest methodology your team can follow on client work, so findings are consistent across testers.

Ramp without a project

Training runs against our targets, so nobody is waiting on a client environment or a signed scope to learn.

Instructor-Led Training

When a whole practice needs to move at once

Offensive track4 days · 4 hrs/day

Your testers work the full OWASP LLM Top 10 against live targets, taught by the founder. Delivered privately to your firm on dates agreed with you — not an open cohort your consultants have to fit around client work.

$999 per seatMinimum 5 seats
Defence track2 days · 4 hrs/day

For firms that also advise on remediation. Consultants build and test real guardrails, scored on security and utility together — so the fix you recommend to a client is one you have proven holds.

$499 per seatMinimum 5 seats

Let's talk about your bench

Tell us how many testers you need AI-ready, and by when. We'll come back with a plan and a price.