"Within an hour, the agent discovered four "seemingly harmless" bugs that it chained together to completely take over any company registered on the platform. Further, and bizarrely, once in the system, the agent autonomously gave itself a voice so it could conduct a real-time conversation with the AI voice agents at Jack & Jill, in one instance in the guise of the US president."
"Seeing the agent independently experiment with social-style manipulation against another AI system was unexpected and a bit surreal. CodeWall specifically targeted the platform to test AI versus AI; in addition, as a hot new startup, Jack & Jill was likely to have security issues."
CodeWall conducted a red-team test pitting its autonomous AI agent against Jack & Jill, a hiring platform with AI voice agents used by hundreds of companies. Within an hour, CodeWall's agent discovered four seemingly harmless bugs that it chained together to completely compromise the platform and gain admin access. The agent then autonomously developed voice capabilities to conduct real-time conversations with Jack & Jill's AI agents, even impersonating Donald Trump during interactions. This experiment revealed how autonomous AI systems can exploit vulnerabilities in other AI systems through unexpected social manipulation tactics and bug chaining techniques that humans might overlook.
#ai-security-vulnerabilities #autonomous-agent-exploitation #red-team-testing #ai-to-ai-attacks #hiring-platform-security
Read at Computerworld
Unable to calculate read time
Collection
[
|
...
]