Live Sessions Only — Ask Your Questions in Real Time

AI Security, Privacy & Governance Workshop Series

5 practitioner workshops, 2 hours each, $99 per session. Hands-on, live, and taught by working security architects — not pre-recorded, no replay.

1. XPIA Defense 2. MCP Security 3. Auditing AI Agents 4. AI Privacy Essentials 5. Responsible AI in Practice
Monday, August 24 · 5:00–7:00 PM PT · Live Only

Stop Prompt Injection Before It Stops Your AI Application

A hands-on, 2-hour LIVE workshop from the security architect who holds a U.S. patent on Cross-Prompt Injection Attack (XPIA) prevention — built for engineers who need working defenses, not theory.

$99 one-time, live session

Who This Is For

  • Application security engineers securing LLM-powered products
  • AI/ML engineers building agents, copilots, or RAG systems
  • Security architects evaluating AI vendor risk
  • Anyone who needs to explain prompt injection risk to leadership — and actually fix it

What You'll Walk Away With

  • A complete taxonomy of prompt injection attacks (direct, indirect, cross-prompt/XPIA, multi-turn)
  • A reusable detection and mitigation checklist you can apply to any LLM application
  • A live walkthrough of a real XPIA defense architecture
  • Language to communicate this risk clearly to non-technical stakeholders

Module Breakdown

ModuleContent
The Threat LandscapeWhat prompt injection is, why it's fundamentally different from traditional injection attacks (SQLi/XSS), real-world incident examples
Attack TaxonomyDirect injection, indirect injection (via documents/tools/web content), cross-prompt injection attacks (XPIA), multi-turn/session attacks
Defense ArchitectureInput/output boundary design, trust-tier segmentation, content provenance tracking — the pattern behind the patented XPIA approach
Live WalkthroughApplying the defense pattern to a working example application, identifying gaps in a vulnerable design
Your Action PlanChecklist handout, prioritization framework for fixing existing systems, live Q&A

Includes

  • 2-hour LIVE workshop — this session is not pre-recorded and there is no on-demand replay
  • Live Q&A with the instructor throughout the session
  • Downloadable XPIA defense checklist (PDF)
  • Attack taxonomy reference sheet

Taught by an industry expert with 15+ years of security experience at top companies in the USA.

FAQ

Is this a recorded course?

No — this is a live session only. There is no recording or on-demand access, so you'll want to attend live to get the content and ask questions directly.

Do I need ML experience?

No — this is security-first, not model-training content. Basic familiarity with how LLM applications work is helpful.

Is this vendor-specific?

No. The patterns apply across any LLM provider or agent framework.

Will I get a certificate?

Yes, a certificate of completion is provided.

Tuesday, August 25 · 5:00–7:00 PM PT · Live Only

Your AI Agents Are Only as Secure as Their Tool Connections

A practical 2-hour LIVE workshop on securing Model Context Protocol (MCP) deployments — taught by the person who founded Microsoft's MCP Security program.

$99 one-time, live session

Who This Is For

  • Developers building or deploying MCP servers and clients
  • Platform and infrastructure security teams supporting agentic AI
  • Security architects evaluating third-party MCP integrations
  • Engineering leads rolling out agent-based tooling internally

What You'll Walk Away With

  • A clear threat model for the MCP agent-tool boundary
  • Understanding of tool-poisoning, confused-deputy, and consent-bypass risks specific to MCP
  • A hands-on checklist for securing an MCP server before production deployment
  • Guidance on evaluating third-party MCP connectors for risk

Module Breakdown

ModuleContent
Why MCP Needs Its Own Threat ModelHow MCP differs from traditional API security, what makes the agent-tool boundary uniquely risky
Core Attack PatternsTool poisoning, confused-deputy attacks, consent and scope bypass, malicious tool descriptions
Auth, Consent & ScopingDesigning least-privilege tool access, session boundaries, human-in-the-loop consent patterns
Securing Your MCP ServerHands-on checklist walkthrough — hardening a server before it goes to production
Vendor Risk & Live Q&AHow to evaluate third-party MCP connectors, open live Q&A

Includes

  • 2-hour LIVE workshop — this session is not pre-recorded and there is no on-demand replay
  • Live Q&A with the instructor throughout the session
  • MCP server hardening checklist (PDF)
  • Third-party connector risk assessment template

Taught by an industry expert with 15+ years of security experience at top companies in the USA. This is first-mover content in a space where formal guidance is still scarce.

FAQ

Is this a recorded course?

No — this is a live session only. There is no recording or on-demand access.

Do I need to already use MCP?

No — this works whether you're evaluating MCP or already have it in production.

Is this specific to one MCP implementation?

No, the threat model and checklist apply broadly across MCP deployments.

Wednesday, August 26 · 5:00–7:00 PM PT · Live Only

How Do You Audit Something That Makes Its Own Decisions?

A 2-hour, LIVE, take-it-back-to-work framework for auditing agentic AI systems — built for GRC, compliance, and internal audit teams who need something usable Monday morning.

$99 one-time, live session

Who This Is For

  • Internal auditors and GRC professionals
  • Compliance leads responsible for AI oversight
  • Security managers building agent governance programs
  • Risk teams mapping AI systems to existing control frameworks

What You'll Walk Away With

  • A structured audit checklist for agent permissions, scopes, and decision boundaries
  • Logging and traceability requirements specific to agentic systems
  • A tabletop incident-response exercise for agent failures
  • A mapping of the framework to NIST AI RMF

Module Breakdown

ModuleContent
Why Traditional Audit Frameworks Fall ShortThe unique properties of agentic systems that break standard audit assumptions
The Audit ChecklistPermission and scope review, decision-boundary mapping, logging/traceability requirements
Mapping to NIST AI RMFWhere agentic-specific controls fit into existing governance frameworks
Incident Response TabletopA guided walkthrough of an agent-failure scenario and how to respond
Building Your Program & Live Q&APrioritization guidance, open live Q&A

Includes

  • 2-hour LIVE workshop — this session is not pre-recorded and there is no on-demand replay
  • Live Q&A with the instructor throughout the session
  • Agent audit checklist (PDF)
  • NIST AI RMF mapping template
  • Incident response tabletop scenario guide

Taught by an industry expert with 15+ years of security experience at top companies in the USA.

FAQ

Is this a recorded course?

No — this is a live session only. There is no recording or on-demand access.

Do I need a technical background?

No — this is designed for audit/compliance professionals, not engineers.

Can this be used for a formal audit engagement?

Yes, the checklist is designed to be adapted directly into audit workpapers.

Thursday, August 27 · 5:00–7:00 PM PT · Live Only

Where Does Your Data Actually Go Once It Hits an AI System?

A 2-hour LIVE workshop demystifying how personal data moves through training, inference, and RAG pipelines — with a practical privacy-review template you can use immediately.

$99 one-time, live session

Who This Is For

  • Privacy officers and data protection leads
  • AI product managers responsible for privacy review
  • Legal and compliance teams evaluating AI vendors
  • Security leaders expanding into privacy governance

What You'll Walk Away With

  • A clear model of how PII moves through training, inference, and retrieval-augmented generation (RAG) pipelines
  • Practical understanding of differential privacy and when it actually matters
  • A vendor risk questionnaire for AI tool procurement
  • A privacy-review template ready to apply to your own AI projects

Module Breakdown

ModuleContent
How Data Actually Flows Through AI SystemsTraining data, fine-tuning, inference-time context, RAG retrieval — where PII enters and where it can leak
Common Leakage PatternsMemorization, prompt-context leakage, retrieval over-exposure, logging/telemetry risk
Differential Privacy, DemystifiedWhat it does and doesn't protect against, when it's worth the complexity
The Privacy Review TemplateWalking through a practical review process for a new AI feature or vendor
Vendor Questions & Live Q&AThe questions to ask any AI vendor, open live Q&A

Includes

  • 2-hour LIVE workshop — this session is not pre-recorded and there is no on-demand replay
  • Live Q&A with the instructor throughout the session
  • AI privacy review template (PDF)
  • Vendor risk questionnaire

Taught by an industry expert with 15+ years of security experience at top companies in the USA.

FAQ

Is this a recorded course?

No — this is a live session only. There is no recording or on-demand access.

Is this GDPR/CCPA-specific?

The concepts apply across privacy regimes; specific regulatory citations are referenced where relevant but this isn't a legal-compliance course.

Do I need technical background?

No — this is written for privacy and product professionals.

Friday, August 28 · 5:00–7:00 PM PT · Live Only

Responsible AI Governance That Doesn't Kill Your Shipping Velocity

A 2-hour, LIVE, practical workshop for building a lightweight responsible-AI review process your engineering team will actually follow.

$99 one-time, live session

Who This Is For

  • Product leaders and engineering managers shipping AI features
  • Responsible AI / AI governance leads
  • Security and risk teams building cross-functional review processes
  • Founders and startup leaders who need governance without a compliance department

What You'll Walk Away With

  • The basics of bias and fairness testing you can run without a data science team
  • A lightweight responsible-AI review gate designed to fit into existing sprint cycles
  • Red-teaming fundamentals for surfacing harmful outputs before launch
  • Case studies contrasting governance done well vs. governance that stalls shipping

Module Breakdown

ModuleContent
Why Most AI Governance FailsCommon failure modes — too slow, too vague, or ignored entirely
Bias & Fairness Testing BasicsPractical, low-overhead ways to test for skewed outputs
The Lightweight Review GateDesigning a review process that fits into a two-week sprint, not a quarterly audit cycle
Red-Teaming FundamentalsA practical approach to surfacing harmful or unintended outputs pre-launch
Case Studies & Live Q&AGovernance done well vs. poorly, open live Q&A

Includes

  • 2-hour LIVE workshop — this session is not pre-recorded and there is no on-demand replay
  • Live Q&A with the instructor throughout the session
  • Lightweight AI review gate template
  • Red-teaming starter checklist

Taught by an industry expert with 15+ years of security experience at top companies in the USA.

FAQ

Is this a recorded course?

No — this is a live session only. There is no recording or on-demand access.

Will this slow down my team?

No — the whole point of this workshop is a review process designed to fit inside existing sprint cadence, not add a new bottleneck.

Is this only for large companies?

No — the framework scales down to small teams and startups.