SECURITYJun 7 · 13:59 UTCFOX NEWS

Is Apple Intelligence on your iPhone really secure?

Researchers discovered vulnerabilities in Apple Intelligence's on-device AI, allowing manipulation through prompt injection and Unicode tricks with a 76% success rate. Apple addressed these issues in iOS 26.4 and macOS 26.4, but the attacks could execute without stealing devices or cracking passcodes.

Nexus surfaces and summarizes. The full story lives at the source.

Mentioned

AppleorganizationRSAC ResearchorganizationApple IntelligencetopiciOS 26.4topicmacOS 26.4topicGoogle Geminitopic

Spot something wrong with this article?Report a problem →

Forward this

Related Signal

Adjacent reporting

TECHNOLOGYMay 7 · 19:18 UTC
Mozilla says 271 vulnerabilities found by Mythos have "almost no false positives"
SECURITYMay 7 · 13:00 UTC
'TrustFall' Exposes Claude Code Execution Risk
SECURITYMay 20 · 19:10 UTC
Google publishes exploit code threatening millions of Chromium users
SECURITYMay 7 · 19:36 UTC
Mozilla says 271 vulnerabilities found by Mythos and "almost no false positives"
SECURITYMay 5 · 15:20 UTC
Kaspersky suspects Chinese hackers planted a backdoor into Daemon Tools in ‘widespread’ attack
SECURITYMay 4 · 19:14 UTC
Exploit Cyber-Frenzy Threatens Millions via Critical cPanel Vulnerability