AML.T0000

MITRE ATLAS / Search Open Technical Databases

Adversaries search publicly available research and technical documentation to learn how and where AI is used inside a victim organization, then use that knowledge to pick targets or tailor an attack. Academic papers, preprint servers, and conference talks by the victim's own staff are common sources.

Think of it likeReading a company's published architectural plans at the library before deciding which wall to knock through.

In plain English

Teams publish more about their AI than they realize. Papers, patents, and conference talks name the model architecture, the training approach, and often the people who built it, which is exactly the head start an attacker wants.