Skip to content

Security & Privacy

Evasion attack

Changing something an AI receives during use to trick it into making a mistake.

Example

Researchers alter a test image so an AI puts it in the wrong group.

Why people use it

It examines attempts to trick AI through the information given during use.

What you'll hear

“Could a changed image cause the detector to miss this?”

What this means for you

Test protections against changes an attacker could realistically make.

Can you control it?

No

No direct control. This describes a wider issue, concept or result rather than something you can simply switch on or off in a tool.

Common questions

Does this attack change the original teaching material?
Not necessarily. It can target what the finished system receives while leaving its earlier learning untouched.
Does a successful attack have to look obvious to a person?
No. Some changes can be hard to notice while still affecting the AI.
Can a defense work in one setting and fail in another?
Yes. Camera conditions, access and the attack method can change the result.

Related terms

Still have questions?

Up to 500 characters.

Ask LATHIC about AI. Relevant glossary entries may be included.

Your question, the glossary entries it matches, and a rotating pseudonymous identifier go to Microsoft Azure’s OpenAI service through Vercel AI Gateway to generate an answer. Zero retention and no training are required of the provider, and LATHIC does not save your question or answer. Privacy Notice