Add selected Workbench capabilities through bounded OEM and partner integrations
AI Security Pricing
Pricing follows the defined scope, systems, workflows, evidence requirements, delivery model, and decision the work must support. Start with one bounded review or pilot and expand only when the next phase has a clear owner and value.
NDA, DPA, SOW, rules of engagement, evidence handling, and subprocessors available through the Trust Center.
AI Launch Security Review
Scoped after triage
5-10 business days
Product teams shipping an AI feature, copilot, RAG system, agent, or workflow in the next 30–60 days. First findings in 5 business days. Typical budget: launch, AppSec, product security, red team, customer assurance, or security review.
Primary services
AI Launch Security Review
Common deliverables
Launch Risk Memo, Graph-Backed Abuse Paths, Release Gate Checklist, Defense Breakpoint Fix Backlog
Expert-Led AI Security Program Baseline
Scoped after discovery
1-3 weeks
Teams that need a fast baseline before assessment, red-team, hardening, governance, or buyer review work.
Primary services
Expert-Led AI Security Program Baseline
Common deliverables
Analyst-Reviewed Program Baseline, AI surface inventory, Control Coverage Snapshot, Evidence-gap register
AI Product Security Assessment
Scoped after discovery
2-4 weeks
A deeper follow-on assessment for broader architecture, trust-boundary, RAG, agent, tenant-isolation, product-security, and evidence coverage after launch triage or when the product surface is already larger than a focused launch review.
Primary services
AI Product Security Assessment, AI Guardrails & Evals Review, AI Security Sales Enablement
Common deliverables
AI system inventory, Architecture and trust-boundary review, Abuse-path review, Control gap assessment
Adversarial Testing & Hardening
Scoped after discovery
3-6 weeks
Teams that need reproducible abuse validation, RAG/XPIA testing, agent action testing, guardrail/eval validation, retesting, or post-assessment hardening.
Primary services
AI Red Team & Adversarial Testing, RAG & XPIA Security Review, Agentic Workflow Security & Hardening, AI Guardrails & Evals Review
Common deliverables
Red-team scope and test plan, Findings register with reproduction evidence, Agent permission matrix, RAG authorization review
Program, Governance & Evidence
Scoped after discovery
4-10 weeks or retainer
Teams that need an AI security operating model, release gates, control ownership, evidence cadence, trust-center support, or a follow-on program build after the baseline.
Primary services
Expert-Led AI Security Program Baseline, AI Security Sales Enablement, AI Product Security Assessment
Common deliverables
AI governance operating model, Control ownership matrix, Evidence pack, Control crosswalk
Flexible Expert Support
Time banks are prepaid AI security engineering hours for advisory, review, remediation planning, evidence review, red-team support, governance support, and follow-up after an assessment. They are not a substitute for a scoped assessment when rules of engagement, testing depth, or formal deliverables are required.
TIME BANK · Starter
10hours
$2,090
$209/hr
Most popular
TIME BANK · Team
25hours
$4,975
$199/hr
TIME BANK · Program
50hours
$8,950
$179/hr
Hours never expire. Transferable within your org. Hours are tracked per session with a running balance in your client portal. Explore engagement options
Academy training
Academy plans provide access to field-guide lessons, labs, study cards, skill checks, and role-based training paths for teams learning how to map, test, harden, and evidence AI systems. Or purchase individual labs and domain packages at /pricing/academy-lab-access.
AI SECURITY ACADEMY · ACCESS
Weekly
$9/week
Auto-renews weekly · Single seat
View planMost popular
AI SECURITY ACADEMY · ACCESS
Monthly
$25/month
Auto-renews monthly · Single seat
View planAI SECURITY ACADEMY · ACCESS
Annual
$150/year
Equivalent to $12.50/mo · Auto-renews annually · Single seat
View planAcademy training does not certify an organization’s AI system as secure and does not replace assessment, red-team testing, or audit work. Team plans (10 seats) from $499/month or $4,990/year. Enterprise (50 seats) at $19,900/year. View all plans
Enterprise course
Private-offer courses for the people who build, test, sell, govern, and operate AI products. Each includes a public-safe manuscript, print edition, and enterprise delivery path.
Academy Course
Secure Coding with GenAI
Private offer
Build, review, test, and ship AI-assisted code safely.
Academy Course
AI Security for Sales Engineers
Private offer
Speak clearly about AI risk, controls, and buyer trust.
Academy Course
AI Red Teaming for Product Teams
Private offer
Test LLM, RAG, and agentic systems before users and attackers do.
Academy Course
AI Product Management for Secure AI Features
Private offer
Define, scope, and ship AI products buyers can trust.
Academy Course
AI Model Gateways and Secure Platforms
Private offer
Build the approved path for safe AI adoption.
Academy Course
Hiring AI-Savvy Talent Without Unicorn Hunting
Private offer
Build better reqs, sharper scorecards, and roles the market can fill.
MADE Domain Packages
Each package includes modules, guided labs, study cards, skill checks, and evidence-writing prompts where implemented. $399 per package — or get all four in Academy Complete at $1,199.
Map Pillar
Map Domain Package
$399
Inventory your AI surface before it becomes an incident.
Attack Pillar
Attack Domain Package
$399
Train teams to find AI failures before they reach production.
Defend Pillar
Defend Domain Package
$399
Turn AI security findings into deployable controls.
Evidence Pillar
Evidence Domain Package
$399
Package AI security work so auditors and governance teams can act on it.
Academy Complete — $1,199 · All four domain packages. Modules, labs, skill checks, flashcards, and evidence-writing prompts across Map, Attack, Defend, and Evidence.
Buy completeDelivery Add-Ons
Add LMS/SCORM setup, Slack or Teams rollout, enterprise onboarding, or evidence review to any Academy or domain package purchase.
LMS / SCORM Setup
$1,500
SCORM/Moodle-ready delivery setup, import guidance, and LMS administrator onboarding.
InquireSlack / Teams Rollout
$2,500
Slack or Teams delivery install, pilot workflow, and onboarding contact coordination.
InquireEnterprise Onboarding
$5,000
White-glove rollout planning with IT/security contact routing and first cohort support.
InquireEvidence Review
$1,500
Validate and polish Academy lab outputs into review-ready AI security evidence summaries.
InquireSoftware & tooling licenses
AI Security Workbench instruments support assessments, red-team work, hardening, and evidence production. Some tools can be licensed separately; others remain engagement-backed unless scoped as a private deployment.
Direct Product
Scoped after discovery
Use Code Scanner directly inside an AI Security LLC engagement or direct product workflow. Headless execution and machine-readable outputs are supported in the scoped deployment mode.
OEM / White-Label
From $50k
Embed Code Scanner behind your existing scanner, AppSec platform, MSSP workflow, or security product through the OEM Engine. 30-day pilot pilot; partner-controlled UX and reporting; JSON and SARIF outputs.
Enterprise OEM
Custom
Deep white-label licensing, custom naming, report language, packaging, and partner-owned presentation where strategic deployment needs more control than a pilot.
AI Code Mini-Scan
Scoped after discovery
Find the highest-priority code-risk paths in one private repo. Attack Path Analysis summary, SARIF, Markdown, and validation plan. Delivered within 5 business days where scope is complete.
AI Product Security Review Sprint
Scoped after discovery
1 app/package, Code Scanner run, product-readiness report, top blockers, remediation checklist, and one review call.
AI Code Attack Path Analysis
Scoped after discovery
1-2 repos, manual triage, Attack Path Analysis Record, validation plan, candidate disclosure register where applicable, Jira backlog, review-ready evidence package, and readout workshop.
How to choose
We do not know where to start
AI Security Program Baseline - Scoped after discovery
We need to map our AI system
AI Product Security Assessment - Scoped after discovery
We are launching an AI feature
AI Launch Security Review - Scoped after triage
We need deeper coverage after launch triage
AI Product Security Assessment - Scoped after discovery
We need abuse validation
AI Red Team & Adversarial Testing - Scoped after discovery
Our agent can take actions
Agentic Workflow Security & Hardening - Scoped after discovery
RAG or tenant data could leak
AI Red Team & Adversarial Testing - Scoped after discovery
Enterprise buyers are blocking deals
AI Security Sales Enablement - Scoped after discovery
Governance is policy, not operations
AI Governance & Security Program Build - Scoped after discovery
We need ongoing support
Time bank, retainer, or phased program - Scoped after discovery
What teams actually scope
AI Feature Launch Review
Customer-facing LLM feature: architecture review, trust-boundary map, abuse-path review, control gap assessment, remediation roadmap.
Scoped after triage
RAG & XPIA Security Review
Knowledge workflow: document ingestion and retrieval review, tenant/permission testing, external-content prompt injection testing, RAG authorization review.
Scoped after discovery
Agentic Workflow Security & Hardening
Agent workflow: tool inventory, permission matrix, approval bypass testing, logging/rollback review, hardening backlog.
Scoped after discovery
Enterprise Security Review Evidence Pack
Questionnaire answer bank, model/provider boundary statement, trust-center AI security notes, evidence references, deal-blocker priorities.
Scoped after discovery
AI Security Program Baseline
Inventory model, control ownership, framework crosswalk, evidence lifecycle, release gate, executive roadmap.
Scoped after discovery
AWS-compatible scope
Common AWS-compatible surfaces we can review include Amazon Bedrock, Bedrock Agents, Bedrock Knowledge Bases, Bedrock Guardrails, SageMaker, Lambda, API Gateway, S3, KMS, Secrets Manager, IAM, CloudTrail, CloudWatch, OpenSearch, Step Functions, EventBridge, Security Hub, GuardDuty, AWS Config, WAF, DynamoDB, RDS/Aurora, SQS/SNS, and CI/CD integrations where relevant.
Current commercial motion is direct consulting/SOW. Marketplace and private-offer packaging can be scoped when relevant.
Workbench-backed delivery
Pricing includes use of Workbench instruments where relevant: Threat Canvas, Application Surface Discovery, Code Scanner, Runtime Trace, Adversarial Range, Authority Graph, RAG Security Testing, AI Control Crosswalk, and Evidence System.
Code Scanner is also available as a separate tooling license for teams that need repeatable, graph-backed code-risk analysis, developer exports, and review-ready evidence. Other Workbench instruments remain engagement-backed unless separately scoped.
Add-ons
FAQ
No. The ranges are public so buyers can budget early. The SOW is fixed after scoping the system, evidence needs, testing depth, and delivery timeline.
Yes. The AI Security Scorecard is the no-cost evaluation entry point. The AI Security Program Baseline is the smallest paid entry path and routes into the right next engagement.
The first fit/scoping conversation is free. Paid work starts when there is a scoped benchmark, assessment, sprint, or program SOW.
No. The scorecard is the evaluation step; the paid offer is the AI Security Program Baseline.
Sometimes. If the larger engagement begins immediately and reuses the same evidence request and scope, we can structure the SOW that way.
Yes. Mutual NDA, SOW, DPA when needed, assessment terms, evidence handling, and red-team rules of engagement can be provided.
Yes, as AWS-compatible architecture review and procurement support where relevant.
No. We provide evidence, mappings, findings, and readiness artifacts. Formal certification or audit opinions require the appropriate auditor or legal authority.
Yes. That is part of AI Security Sales Enablement when scoped.
Yes. Retesting can be included or added as a separate package.
Implementation support is not included by default. It can be scoped as hardening, release-gate, or advisory support.
Both. Direct product use can be scoped for direct workflows, and OEM / white-label licensing is available for partner products through the OEM Engine. Workbench-backed assessments and review sprints add expert review, triage, and reporting. Separate private repo delivery, CI integration, and enterprise deployment can be scoped.
No. It helps teams prepare evidence, identify blockers, and remediate AI-native security risks before official marketplace, enterprise buyer, or partner security review.
No. It ranks CVE candidates and can generate disclosure-ready evidence when appropriate, but CVE assignment depends on affected products, validation, vendor coordination, public references, and the appropriate CNA or advisory process.
Bring the AI product, workflow, buyer blocker, launch risk, or governance gap. The first scoping step determines which MADE path and pricing lane fits.
Pricing packet
Self-serve checkout, quote-first services, and invoice-first retainers all attach the relevant contract docs and fulfillment metadata.
Mutual NDA
Mutual confidentiality protections for pre-sales, delivery, and research collaboration contexts.
Used for NDA baseline
Statement of Work Template
Mission-specific scope, deliverables, timeline, access, assumptions, and acceptance criteria for scoped AI security engagements.
Assessment Terms Addendum
Scope, authorization, evidence use, testing boundaries, safe harbor, retesting, reporting limitations, and reliance limits for AI product security assessments.
Used for Assessment and quote-first services
Data Processing Addendum
Controller/processor allocation, data protection obligations, subprocessing, security measures, AI provider boundaries, and customer-data handling for scoped services.
Used for Data handling and public-safe claims
Commercial Services Addendum
Converts the services framework into scoped paid work with rate card, invoicing, and activation terms.
Used for Retainer and advisory billing
Publication & Claim-Readiness Policy
Claim-readiness criteria for public research, trust pages, scorecards, attestations, sponsor materials, security review outputs, and buyer-facing evidence.
Used for Data handling and public-safe claims · Sponsored research and launch
Payment terms
Prepaid for self-serve items, quote-first for services, and invoice-first for retainers.
Acceptance terms
Accepted on delivery or milestone completion depending on the path.
Fulfillment workflow
pricing -> packet -> checkout or quote -> webhook -> launch room