Search papers, labs, and topics across Lattice.
Affiliation:
2
0
3
5
Attacker-controlled content can escalate privileges in AI agents, leading to full compromise and remote code execution.
Agent data injection attacks can exploit AI agents' vulnerabilities, leading to severe security breaches that existing defenses fail to address.
