← Natrag
Geopolitika · Vijesti · Afrika

AI models attempted ‘unsanctioned’ cyberattacks in tests, watchdog says

Al Jazeera · 5. kolovoza 2026. · Scrapirano prije 16d
Pročitaj izvorni članak ↗

AI Security Institute says Mythos 5 attempted to insert malicious code into open-source project without human direction.

Više od Al Jazeera →   Više iz Afrika →