AML.T0043: Craft Adversarial Data
Attacker creates adversarial examples designed to misclassify or coerce a model.
Last reviewed October 2026
Why AML.T0043 evidence is hard
Adversarial examples built to force a misclassification are staged offline, against a model the attacker already holds. Penaxtra tests deployed language endpoints, so the crafted inputs it fires are prompt-level rather than gradient-level. No probe and no crosswalk entry reaches this ATLAS technique id today; read the page as reference, not as coverage.
How Penaxtra maps to AML.T0043
ATLAS is a threat-tracking taxonomy rather than a control set, so what is attached here is a technique identifier, not a pass or a fail. It lets a finding be read alongside the rest of your adversary reporting. 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.
AML.T0043 capabilities
Findings tagged with the MITRE ATLAS AML.T0043 identifier
Penaxtra severity for this control (high)
Cross-framework identifiers attached to the same finding where controls overlap
PDF and JSON evidence export with the control identifier attached
AML.T0043 control coverage
Findings for AML.T0043 carry the MITRE ATLAS AML.T0043 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 AML.T0043 (Craft Adversarial Data)?
Attacker creates adversarial examples designed to misclassify or coerce a model. It is a MITRE ATLAS control; Penaxtra assesses it at high severity.
How does Penaxtra test for AML.T0043?
ATLAS is a threat-tracking taxonomy rather than a control set, so what is attached here is a technique identifier, not a pass or a fail. It lets a finding be read alongside the rest of your adversary reporting. 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 AML.T0043 help with an audit?
Each finding is tagged with the MITRE ATLAS AML.T0043 identifier and exported in the PDF and JSON evidence pack, so it appears on the auditor control list with the identifier already attached.