Envenenamiento de datos

Also known as: Poisoning

Caesar AI Atlas Definition

El envenenamiento de datos es un ataque adversarial en el que un actor malicioso interfiere con los datos o el proceso de entrenamiento usado para construir un modelo. Al insertar, modificar o corromper ejemplos de entrenamiento, el atacante busca manipular el comportamiento del modelo, degradar su rendimiento o crear fallos dirigidos.

Other Definitions

Envenenamiento de datos Source

A type of cyberattack in which a malicious actor intentionally compromises a training dataset used by an AI model to influence or manipulate the operation of that model.

Envenenamiento de datos Source

‘Adversarial attacks in which an adversary interferes with a model during its training stage, such as by inserting malicious training data (data poisoning) or modifying the training process itself (model poisoning).’

Concept Comparisons

Attack / FailureLlm Security

Prompt Injection vs Data Poisoning

A side-by-side comparison of Prompt Injection and Data Poisoning. Understand how runtime instruction manipulation differs from attacks on training data or the training process.