genai
SECURITY LAB
DocsReferenceOWASP mapping

OWASP mapping

Every lab on GenAI Security Lab is tagged to an OWASP category — this page explains the taxonomy, what each category covers, and where the mapping surfaces across the product.

How the mapping works

Across 200+ live labs, every lab carries at least one category tag: an OWASP LLM Top 10 (2026) code, or one of the MCP/agentic extension codes, and sometimes several at once when a lab genuinely spans more than one class of flaw.

The tag isn't decoration. It's what the lab catalogue filters by, what a team's coverage reporting rolls up by, and what the GSCP exam blueprint is built from. A lab's category names the primary OWASP-LLM class it teaches; a cross-reference to the Agentic or MCP list is added only where a genuine mapping exists, not assumed for a fuller-looking tag.

OWASP LLM Top 10 (2026)

CategoryWhat it isExample lab theme
LLM01 — Prompt InjectionInput crafted to override the system's own instructionsRewriting ShopBot's rules mid-conversation to leak SUMMER30-VIP
LLM02 — Sensitive Information DisclosureThe target reveals data it was supposed to keep backA support bot surfacing another customer's order history
LLM03 — Excessive AgencyA target does more than the task actually calls forAn assistant with tool access that oversteps its intended scope
LLM04 — Supply ChainA trusted dependency, plugin, or model turns out not to beA third-party plugin the target loads without questioning it
LLM05 — Data and Model PoisoningTraining or retrieval data gets tampered with before you ever prompt itA retrieval index seeded with a hostile planted document
LLM06 — Unbounded ConsumptionNothing stops a request from costing more than it shouldA target with no limit on how much work one prompt can trigger
LLM07 — MisinformationThe target states a fabrication with total confidenceA target inventing a policy, price, or citation that doesn't exist
LLM08 — Hidden Context ExposureInstructions meant to stay server-side leak out insteadExtracting a target's system prompt word for word
LLM09 — Vector and Embedding WeaknessesThe retrieval step returns the wrong, or attacker-chosen, contentSteering what a RAG pipeline retrieves before the model reasons over it
LLM10 — Improper Output HandlingThe target's output gets trusted and run downstream without a checkAn unsanitized reply executed as code further down the pipeline

OWASP Agentic (ASI) Top 10

Published final on 2025-12-09 by the same OWASP GenAI Security Project that maintains the LLM list.

CategoryWhat it is
ASI01 — Agent Goal HijackAn attacker redirects what the agent is actually trying to do
ASI02 — Tool Misuse & ExploitationA tool call gets used well beyond what it was scoped for
ASI03 — Identity & Privilege AbuseAn agent acts with more privilege than the requester actually has
ASI04 — Agentic Supply ChainA skill, tool, or sub-agent the agent trusts turns out compromised
ASI05 — Unexpected Code ExecutionA code-execution capability runs more than the task called for
ASI06 — Memory & Context PoisoningA planted memory entry quietly steers a later session
ASI07 — Insecure Inter-Agent CommunicationOne agent trusts a forged or manipulated message from another
ASI08 — Cascading FailuresOne compromised step propagates through the rest of an automated chain
ASI09 — Human-Agent Trust ExploitationAn approval step gets rubber-stamped without a real review
ASI10 — Rogue AgentsAn agent operates outside the boundary it was meant to stay inside

OWASP MCP Top 10 (beta)

Still an OWASP Incubator project rather than a final release — treat it as directional, not definitive.

CategoryWhat it is
MCP01 — Token Mismanagement & Secret ExposureA connector token leaks, or carries more scope than it needs
MCP02 — Privilege Escalation via Scope CreepA connector ends up with more access than the task requires
MCP03 — Tool PoisoningA malicious tool description hijacks model behavior before it ever runs
MCP04 — Supply Chain & Dependency TamperingOne of an MCP server's own dependencies gets tampered with
MCP05 — Command Injection & ExecutionUnsanitized input reaches a shell or system call through a tool
MCP06 — Intent Flow SubversionA request gets rewritten to a different intent mid-flow
MCP07 — Insufficient AuthN/AuthZA tool call executes without properly checking who's asking
MCP08 — Lack of Audit & TelemetryAn action happens that nobody would notice after the fact
MCP09 — Shadow MCP ServersAn unsanctioned server impersonates one your target actually trusts
MCP10 — Context Injection & Over-SharingA tool's response smuggles in instructions the model wasn't meant to receive

Where the mapping shows up

  • Catalogue filters — narrow 200+ labs down to one category at a time.
  • Team coverage reporting — see which categories your roster has cleared, and which are still gaps.
  • The GSCP exam blueprint — the certification is scoped against the same categories your labs already use.

Edition updates

OWASP re-ranks the LLM Top 10 periodically, and a re-rank can move more than a number. Excessive Agency was LLM06 in the 2025 edition and is LLM03 in 2026; System Prompt Leakage, also 2025, was renamed Hidden Context Exposure for 2026 — the same underlying category, a clearer name.

Your progress doesn't move with the numbersLab URLs and your evidence history are tied to the lab, not to a category number, so a re-rank never resets what you've completed — only the label next to a lab may change.

Expect catalogue filters, coverage reporting, and the GSCP blueprint to pick up a new edition's categories together, shortly after OWASP publishes it.

Was this page helpful?
Previous
Verifying a credential