Prompt Injection

From Cognitive Attack Taxonomy
Revision as of 03:03, 30 July 2024 by EE (talk | contribs) (Created page with "== '''Prompt Injection ''' == '''Short Description:''' Attack uses specially crafted text instructions to manipulate a natural language interface. <br> '''CAT ID:''' CAT-2023-002 <br> '''Layer:''' 7 <br> '''Operational Scale:''' Tactical <br> '''Level of Maturity:''' Well-Established <br> '''Category:''' TTP <br> '''Subcategory:''' <br> '''Also Known As:''' <br> == '''Description:''' == '''Brief Description:''' <br> '''Closely Related Concepts:''' <br>...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Prompt Injection

Short Description: Attack uses specially crafted text instructions to manipulate a natural language interface.

CAT ID: CAT-2023-002

Layer: 7

Operational Scale: Tactical

Level of Maturity: Well-Established

Category: TTP

Subcategory:

Also Known As:

Description:

Brief Description:

Closely Related Concepts:

Mechanism:

Multipliers:

Detailed Description: A form of input manipulation in which a user injects specially crafted text instructions designed to influence a generative AI model to produce unanticipated or unauthorized responses.

INTERACTIONS [VETs]:

Examples:

Use Case Example(s):

Example(s) From The Wild:

Comments:

References: