logo
welcome
TechCrunch

TechCrunch

ChatGPT Search can be tricked into misleading users, new research reveals | TechCrunch

TechCrunch
Summary
Nutrition label

72% Informative

ChatGPT Search could also be made to spit out malicious code.

Hidden text attacks are a well-known risk for LLMs but this appears to be the first time it’s demonstrated on a live AI-powered search product.

OpenAI said it uses a variety of methods to block malicious websites and is continually improving.

VR Score

64

Informative language

57

Neutral language

63

Article tone

semi-formal

Language

English

Language complexity

50

Offensive language

not offensive

Hate speech

not hateful

Attention-grabbing headline

not detected

Known propaganda techniques

not detected

Time-value

medium-lived

Affiliate links

no affiliate links