- — Shadow AI everywhere, unowned and untagged
- — Company data pasted into public chatbots
- — No threat model defined for the model, data, or agent.
- — Staff unprepared for deepfakes and voice clones
- — No policy, no rules of the road
Applied Security Credential · Issued by GAISB
GAISB AI Systems Security Architect™
Understand the whole picture of AI security — before you implement. See every surface, every threat, every rule, and carry a documented AI Security Blueprint into the decision.
An evergreen, strategic credential: concepts and judgment, not tools or code. Nothing in it expires when the next model ships.
9 modulesEach ends in a guided lab
EvergreenConcepts & judgment, not tools
3–6 wksTypical completion
Enrollment OpenJoin the next live cohort
Registry-verified credentialregistry.gaisb.org · LinkedIn-ready badge
The AI Security Blueprint Method™ · Map the risk. Build the defense. Govern the use.
NIST AI RMFEU AI ActISO/IEC 42001OWASP
The operational reality
Organizations do not need more AI tools. They need to understand the risk before they turn AI loose.
AI is already inside most organizations — pasted into chatbots, wired into workflows, adopted team by team with no threat model behind it. Classical security guarded the doors against outsiders. This new intelligence works on the inside.
This credential trains professionals to see the whole board — every surface, every threat, every rule — and to carry a documented AI Security Blueprint into the implementation decision.
- ✓ Exposure mapped across the organization
- ✓ Threats named, classified, and prioritized
- ✓ The human layer briefed and trained
- ✓ Data and IP exposure protected
- ✓ The rules understood before they bind
The mental model
Imagine AI as a brilliant new hire with genius knowledge and zero judgment.
It has read the whole internet, works at superhuman speed, has no memory of your rules unless you install them, and will do almost anything a confident voice tells it to. Everything confusing about AI security becomes obvious once you picture the person you just hired.
The new hire · Eight ways it betrays you in thought
T-01The Whisper — Injection
T-02Poisoned Data
T-03The Leak
T-04Deepfakes
T-05IP Exposure
T-06Rogue Actions
T-07Supply Chain
T-08Shadow AI
The three surfaces · Every AI can be attacked throughSurface / 01 Surface / 02 Surface / 03
Model
The brain. What it believes, and who can talk it into anything.
Data
The memory. What it has read, and what it might repeat.
Agent
The hands. What it can touch, and what happens when it acts.
The durable questions: What can it touch? Who can influence it? What happens when it gets it wrong? Tools change constantly; these surfaces and questions do not.
The Method
Map. Defend. Govern.
The AI Security Blueprint Method™ — the whole method in three moves. Every lab in the program expands from these three words.
Map
See where AI touches the organization — its surfaces, its threats, its blast radius.
Defend
Apply durable principles and defensive AI to shrink the risk before it lands.
Govern
Align to the rules, document the decisions, and keep the blueprint alive.
The AI Systems Security Architect
From using AI to seeing every way it can be attacked.
The credential develops the whole map across nine modules. Every module ends in a guided lab — run in Claude or ChatGPT, no tools, no code — that produces one layer of the learner's own AI Security Blueprint.
Module 01Why AI Breaks Classical SecurityWhy guarding the doors no longer works when the intelligence is on the inside.
Module 02The Three SurfacesEvery system decomposed into model, data, and agent.
Module 03The Attacker's PlaybookThe six durable attack classes, ranked for your context.
Module 04The Human LayerDeepfakes, voice clones, and phishing at scale — briefed plainly.
Module 05Your Data, Your IP, Your ExposureWhere prompts leak copyright, trade-secret and privacy exposure.
Module 06The Defender's MindsetLeast privilege, human-in-the-loop, kill switch, defense in depth.
Module 07Using AI to Secure AIWhere defensive AI earns its place: monitoring, detection, red teaming.
Module 08The Rules of the RoadNIST AI RMF, EU AI Act, ISO/IEC 42001, OWASP — what binds you.
Module 09Capstone — The Blueprint AssembledAll layers assembled, plus the pre-implementation decision checklist.
This is a strategic, evergreen credential — concepts and judgment, not tools or code. Nothing in it expires when the next model ships.
“Capability is demonstrated by a blueprint that maps, defends, and governs an organization’s AI — not by the number of tools a candidate can name.”
GAISB Assessment PrincipleProfessional capability
What a GAISB AI Systems Security Architect™ can demonstrate.
The credential is defined by demonstrable capability, not tool coverage. Architects are expected to see the whole board and prove it.
01
Map an organization's AI exposure.
02
Decompose any AI system into model, data, and agent.
03
Classify the six durable attack types.
04
Brief a team on AI-powered social engineering.
05
Audit data and IP exposure.
06
Apply the six defense principles.
07
Identify defensive AI use cases.
08
Scope which AI regulations apply.
09
Assemble an organizational AI Security Blueprint.
10
Run the pre-implementation checklist.
Professional proof of work
The credential requires a completed AI Security Blueprint.
Every candidate assembles and submits one AI Security Blueprint, built layer by layer through the nine labs, meeting the published GAISB standard before the credential can be issued.
Certification examination
Total questions60
Duration90 minutes
FormatOnline, governed
Passing score70%
Credential renewalEvery 2 years
The AI Security Blueprint
Blueprint layersNine
ReviewReviewed submission
Recorded walkthrough8–12 min
Review areasExposure AccuracyThreat PrioritizationHuman-Layer ClarityData & IP RigorDefense-AI JudgmentGovernance FitPresentation Quality
What you earn
An applied credential designed for public verification.
Applied Security Credential · This certifies that[ Candidate Name ]
has earned the GAISB AI Systems Security Architect™ credentialHaving satisfied the education, applied-blueprint and assessment requirements established for the credential.
✓
Verifiable digital credential — a public record in the GAISB registry at registry.gaisb.org.
✓
LinkedIn-ready badge — add the verified credential to your professional profile.
✓
Reviewed blueprint on record — your assembled AI Security Blueprint, assessed against the published standard.
✓
Evergreen posture — a strategic credential of concepts and judgment that does not expire when the next model ships.
✓
What partners can verify — credential title, level, status, issue date and renewal.
The professional pathway
Where this credential sits.
The security specialization of the GAISB credential family — a sibling track to Governance, built on foundation and professional-core capability.
GAISB AI Essentials™
Practical, provable foundational AI fluency for work, study, and everyday life.
GAISB AI Practitioner™
Professional core AI knowledge and judgment, verified through a governed examination.
GAISB AI Systems Security Architect™
See the whole board of AI security — before the organization implements.
Admission & enrollment
Your path to the Security Architect credential.
$3,995USD
Certification examination voucher — $395 USD, purchased separately. Valid 12 months from issue; one attempt per voucher. A new voucher is required for each retake.
✓ Nine-module curriculum✓ Nine guided AI labs✓ Live cohort sessions✓ Assessment preparation✓ AI Security Blueprint (reviewed)✓ Verifiable credential + badge
Enroll in the Program — $3,995The question you're really asking
Is this worth it?
Answered honestly.
Is this a legitimate certification?
Yes. GAISB AI Systems Security Architect™ is a private professional certification issued by GAISB™, the Global AI Standards Body — a private professional standards organization.
Do I need a security or coding background?
No. GAISB AI Essentials™ is recommended but not mandatory. The credential develops strategic concepts and judgment — not tool operation or code.
Is this a hands-on, tools-based course?
No. Every module ends in a guided lab run in Claude or ChatGPT, with no additional tools and no code. The output is your own AI Security Blueprint — the artifact reviewed for the credential.
What is the AI Security Blueprint?
A documented map of an organization's AI exposure, threats, defenses, human-layer readiness, data and IP posture, and governance fit — assembled layer by layer through the nine labs and submitted as the applied artifact.
How is it assessed?
A governed 60-question, 90-minute knowledge check at 70% to pass, plus a reviewed submission of the assembled AI Security Blueprint with an 8–12 minute recorded walkthrough.
What does it cost and how do I enroll?
Program tuition and the certification examination voucher are confirmed at checkout via Prompt Atlas — GAISB's official enrollment partner — and through official GAISB channels.
Where does it sit in the GAISB pathway?
It is the security specialization of the GAISB credential family — a sibling track to Governance, built on foundation and professional-core capability.
See the whole board before you implement.
Map the risk. Build the defense. Govern the use.
Issued by GAISB — the Global AI Standards Body · Secure enrollment via Prompt Atlas · Verify at registry.gaisb.org