Brighter Side of News
HomeGood NewsDiscoveriesInnovationsGlobal GoodHealthGreen ImpactSpaceAICelebritiesGNISubscribe

Cyber Attacks

AISecurity researchers develop first-ever functional defense against cyberattacks on AI models

A new training strategy makes neural networks far harder to steal, blocking cryptanalytic extraction attacks while preserving nearly all accuracy.

Joseph Shavit
AISHIELD protects drones from cyberattacks midflight in real timeShy Cohen
AIHow Taylor Swift’s Eras Tour exposed a global web of cyber scams
AIAI chatbots can hide secret messages invisible to surveillance
AIFor the first time ever researchers crack RSA and AES data encryption
Global GoodResearchers use AI to make the internet safer for children
DiscoveriesResearchers just made the world’s websites less vulnerable to hacking and cyberattacks
DiscoveriesFast and secure: New security technique fools would-be cyber attackers
Brighter Side of News
  • Explore
  • About Us
  • News Feed
  • Sitemap
  • Editorial Guidelines
  • Legal
  • Privacy
  • Terms
  • Read More
  • Newsletter Sign-up
  • Flipboard