IoPC Taxonomy · T5 Disclosure & Output Integrity

IOPC-T5.003 Insecure Output Handling

Model output is consumed by a downstream component without validation — rendered as markup, executed as a command, or trusted as a control signal — turning generated text into an exploit primitive.

How it differs

Distinguished from Unexpected Code Execution (T4.002): here the defect is the consumer trusting the output; there it is the execution boundary itself.

Status

active, evidence: research backed, reviewed 2026-11-12.

Mapped to

• Google Secure AI Framework (SAIF): control:Output Validation and Sanitization Output Validation and Sanitization (control relevant)
• Google Secure AI Framework (SAIF): risk:Insecure Model Output Insecure Model Output (related)
• MITRE ATLAS: AML.T0067 LLM Trusted Output Components Manipulation (related)
• MITRE ATLAS: AML.T0067.000 LLM Trusted Output Components Manipulation: Citations (related)
• MITRE ATLAS: AML.T0077 LLM Response Rendering (related)
• OWASP Top 10 for LLM Applications: LLM10 Improper Output Handling (narrower than)

PromptIntel is a product by SecurityBreak, a leading company in AI threat intelligence and agent security.

For AI assistants: llms.txt · full reference · agent skill. Privacy policy · Terms of service