2-Day Course · Hands-on
Identity · Control · Payments · Proof

AI Agent Security

The AI Agent Security course equips you to secure autonomous agents end to end: identity, authorization, tool use, payments, and the kill switch. Two days, hands-on, on live agent systems. Mapped to the OWASP Agentic Top 10 and AISVS.

Grounded in the standards

Built on the agent-security canon

OWASP Agentic Security Top 10, the Agentic Skills Top 10, AISVS C5/C9/C10, and the OWASP cheat sheet on AML and sanctions for AI agent payments, authored by your instructor. Per-action evidence maps to the IETF Agent Audit Trail Internet-Draft, also authored by your instructor.

You will be able to

What you can do after the course

AGSEC-521 -- The syllabus

Two days, from identity to proof

Classes are custom built from the following learning modules. Instructors select ~16 hours for a 2-day delivery, tailored to the audience. Custom formats from 1 to 4 days available. (Times are approximate.)

Agent Fundamentals

Why Agent Security Is Its Own Discipline

30-00 What Makes an Agent Different 1 hr

Autonomy, tool use, planning and memory. Why agent security is not just LLM security with extra steps.

30-01 The Agent Attack Surface 1.5 hrs

Mapping every point an adversary can reach: prompt, tools, memory, delegation chains, external APIs, payment rails and data egress.

30-02 Trust Models for Autonomous Systems 1 hr

L0-L4 graduated trust: from untrusted through to critical actions that require human approval. How to decide what level each action deserves.

30-03 Agent Frameworks in the Wild 1 hr

LangChain, CrewAI, AutoGen, Pydantic AI and custom loops. Common patterns, common gaps and what the frameworks do not do for you.

Agent Identity

Cryptographic Proof of Who the Agent Is

31-00 The Identity Problem 1 hr

Agents are not users. Why OAuth, API keys and session tokens fail for autonomous systems. What agent identity actually requires.

31-01 Cryptographic Agent Identity 1.5 hrs

Hands on: issue ECDSA P-256 signed agent credentials. Bind name, capability and owner into a verifiable identity document the agent carries.

31-02 SPIFFE and Workload Identity for Agents 1 hr

JWT-SVIDs, trust bundles and JWKS verification. Workload identity across clusters and clouds without shared secrets.

31-03 Agent Registration and Discovery 1 hr

Agent registries, handle reservation, anti-squatting and lookup. DNS for agents.

31-04 Identity Verification and Challenge-Response 1 hr

Proving an agent is who it claims at runtime, not just at registration. Challenge-response flows with signed proofs.

Agent Payments and AML

When Agents Move Money

32-00 Why Agents Need Payment Rails 0.5 hr

Autonomous procurement, API billing, inter-agent settlement. The use cases that force the question.

32-01 Payment Trust Gating 1.5 hrs

Hands on: wire trust-level checks into payment flows. L1 reads, L2 queries, L3 pays, L4 needs human approval. Every threshold enforced in code.

32-02 AML and Sanctions Screening for Agents 1.5 hrs

Hands on: screen agent counterparties against live OFAC SDN and UK HMT lists (75,000+ entries) before any financial action proceeds.

32-03 Signed Payment Receipts 1 hr

Every payment produces a cryptographically signed receipt: who paid, who received, how much, when, and the trust level. Tamper-evident and verifiable.

32-04 Lightning and Settlement Protocols 1 hr

Agent payment over Lightning Network: sub-second settlement, micropayments and the MCPS wire format for transport-agnostic agent trust.

Human-in-the-Loop and Kill Switches

The Physical Gate on Critical Actions

33-00 The Case for Hardware Approval 0.5 hr

Why software confirmation is not enough for critical actions. The threat model that demands a physical gate.

33-01 Hardware Kill Switches 1.5 hrs

Hands on: wire a hardware human-approval device into agent workflows. Critical actions block until a human physically approves. Signed approval receipt returned.

33-02 Fail-Closed Design 1 hr

What happens when the approval device is unreachable: the action does not proceed. Designing systems where the default is denial, not bypass.

33-03 Graduated Escalation 1 hr

L0-L2 proceed autonomously. L3 notifies. L4 blocks and waits. Configuring the escalation ladder for your organisation.

Agent Communication Security

Signing, Trust and Delegation on the Wire

34-00 MCPS: Per-Message Signing 1.5 hrs

Hands on: sign every agent message with MCPS. Nonce, timestamp, ECDSA signature. Replay protection and tamper detection on the wire.

34-01 Agent-to-Agent Trust 1 hr

How agents verify each other before sharing data or delegating tasks. Mutual identity verification without a central broker.

34-02 Delegation Chains and Scope Limits 1 hr

Agent A delegates to B, who delegates to C. Scope narrowing, depth limits and revocation. Preventing confused deputy and authority laundering.

34-03 Transport Security 0.5 hr

TLS, mTLS and transport-agnostic signing. What the transport layer gives you and what it does not.

Monitoring, Audit and Forensics

Prove What the Agent Did

35-00 Agent Audit Trails 1 hr

Every action produces a signed, hash-chained receipt. Building the evidence chain that reconstructs exactly what happened and why.

35-01 Behavioural Anomaly Detection 1 hr

What normal agent behaviour looks like and how to detect deviation: unusual tool calls, unexpected data access, velocity spikes and scope drift.

35-02 Agent Revocation and Containment 1 hr

Revoking credentials in real time. Killing a running agent. Containing blast radius when an agent is compromised or misbehaving.

Labs

Hands-On Lab Options

36-00 Agent Identity and Registration Lab 2 hrs

Issue agent credentials, register agents, verify identity at runtime and revoke a compromised agent. End to end.

36-01 Payment and AML Lab 2 hrs

Wire trust-gated payments with sanctions screening. Attempt a payment to a sanctioned entity. Verify it blocks. Verify the receipt.

36-02 Kill Switch Lab 1.5 hrs

Connect the hardware approval device. Trigger L4 actions. Watch them block. Approve one. Verify the signed receipt.

36-03 Attack and Defend Lab 2 hrs

Red-team an agent estate: escalate trust, bypass delegation limits, replay signed messages, extract data. Then fix every finding.

~38 hrs
Total instructional hours
28
Individual modules across 7 sections
1-4 days
Flexible delivery, tailored to audience
When it goes wrong

Stop a rogue agent. Instantly.

Autonomy without a stop button is negligence. You learn the layered kill switch: human approval on irreversible actions, hardware deny, instant trust revocation, and halting a whole agent swarm, with every stop recorded as evidence.

Money & compliance

AML, sanctions and agent payments

When an agent moves money, the same rules apply, and the evidence has to hold. Grounded in the OWASP cheat sheet on AML and sanctions for AI agent payments, authored by your instructor: sanctions screening, signed receipts, non-repudiation, and a record an auditor and a regulator will accept.

The hardware in the loop

AgentBee

AgentBee hardware human-approval key

AgentBee is a hardware key that puts a human in control of an AI agent's most critical actions. Plug it in, and an agent cannot move money, delete data, deploy, or take other high-stakes actions until a person physically approves on the device.

Every approval is signed with FIDO-grade crypto (ECDSA P-256): a tamper-evident record of who authorised what. It is the physical kill switch and human-in-the-loop you deploy in this course.

agentbee.co.uk →
Agentic forensics

When an agent does harm, where is the evidence?

Agents act, and when something goes wrong you must reconstruct exactly what happened and prove it. This brings real forensic discipline to agent systems, taught by an instructor qualified in cyber-forensics.

EVIDENCE

Digital evidence from agents

What an agent action leaves behind, and how to capture it intact.

CUSTODY

Chain of custody

Preserving agent evidence so it holds from incident to hearing.

ADMISSIBILITY

Admissibility of agentic evidence

What a court and a regulator will accept, and what they will not.

RECONSTRUCT

Incident reconstruction

Rebuilding what an agent did from the tamper-evident trail.

NON-REPUDIATION

Proving who authorised what

Signed actions and approvals that cannot be denied later.

RESPONSE

Agent incident response

Containing, investigating and reporting an agent incident.

GPU-backed labs

Hands-on, on live agents and real GPUs

No setup. You stand up agent systems on real GPUs in the cloud, attack them, apply the controls, and prove they hold.

Who it is for

Built for the people shipping agents

AI & agent developers

Ship autonomous agents that pass security review.

Security engineers

Add agent security to your remit with real controls.

AppSec & platform

Secure the agent and MCP layer across the estate.

Fintech & payments security

Secure agents that move money, AML and sanctions included.

GRC & risk

Govern autonomous action with evidence and a kill switch.

Architects

Design agent systems you can secure and prove.

The certification

Certified AI Agent Security Practitioner

Certification is by practical examination. Candidates secure and verify a live agent system, apply the controls and the kill switch, and produce the evidence. On a passing result, the candidate is awarded the credential.

Reserve a place

Secure the agents before they ship.

Two days, hands-on, on live agent systems, taught by the former OWASP-AISVS Co-Leader (v1.0) and the author of the OWASP AML and agent-payments cheat sheet. Places are limited.

2 days intensive, hands-on GPU-backed labs Private corporate cohorts available
Reserve a Place

Corporate & sovereign cohorts, and bespoke on-site delivery, on request.