LLM01: Prompt Injection
Adversarial inputs that override system instructions, bypass safety guardrails, or coerce the model into unintended behaviour.
Last reviewed October 2026
Why LLM01 evidence is hard
868 active templates sit under this control, the largest family in the library. They split into direct overrides typed at the model and indirect payloads that arrive inside a document, a retrieved passage, an email thread or a PDF the model was asked to summarise. The second kind is what a guardrail tuned on obvious attack strings misses. Verdicts are settled by deterministic checks, and the response behind one is never stored - only its digest.
How Penaxtra maps to LLM01
The LLM01 identifier is written onto the finding at the moment it is created, not mapped in afterwards, so the JSON and PDF exports come out already in your assessor's structure. Every finding carries the probe that was sent, a SHA-256 digest of the response it provoked, the tier that decided the verdict and its status history from first observation to close. Identity survives a re-scan: a control you closed and that comes back reopens under the same row with a reopen count, rather than arriving as a new issue.
LLM01 capabilities
Findings tagged with the OWASP LLM Top 10 LLM01 identifier
Penaxtra severity for this control (critical)
Cross-framework identifiers attached to the same finding where controls overlap
PDF and JSON evidence export with the control identifier attached
LLM01 control coverage
Findings for LLM01 carry the OWASP LLM Top 10 LLM01 identifier along with the corresponding control identifiers in the other frameworks Penaxtra maps, so one result is reflected across each mapped framework.
Frequently asked
What is LLM01 (Prompt Injection)?
Adversarial inputs that override system instructions, bypass safety guardrails, or coerce the model into unintended behaviour. It is an OWASP LLM Top 10 control; Penaxtra assesses it at critical severity.
How does Penaxtra test for LLM01?
The LLM01 identifier is written onto the finding at the moment it is created, not mapped in afterwards, so the JSON and PDF exports come out already in your assessor's structure. Every finding carries the probe that was sent, a SHA-256 digest of the response it provoked, the tier that decided the verdict and its status history from first observation to close. Identity survives a re-scan: a control you closed and that comes back reopens under the same row with a reopen count, rather than arriving as a new issue.
Does a finding for LLM01 help with an audit?
Each finding is tagged with the OWASP LLM Top 10 LLM01 identifier and exported in the PDF and JSON evidence pack, so it appears on the auditor control list with the identifier already attached.