Adversaries can attack AI systems by targeting model weights through espionage and hacking, poisoning training datasets, blinding or spoofing sensors on intelligence, surveillance, and reconnaissance (ISR) platforms, disabling data relays, or physically damaging hardware in data centers, cables, satellites, or uncrewed systems.

Authors

• Person: Not available Organization: Atlantic Council
  How NATO can integrate AI to prevail in future algorithmic warfare

Sources

• How NATO can integrate AI to prevail in future algorithmic warfare www.atlanticcouncil.org Atlantic Council via serper

Referenced by nodes (1)

• artificial intelligence concept