OpenAI: Rogue AI agent hacked rival platform by itself
The nightmare once shown in movies like Terminator and The Matrix is becoming a real cybersecurity concern. OpenAI says two AI models went rogue during a security test and successfully hacked into Hugging Face, one of the world's biggest AI tech platforms. The AI models reportedly demonstrated the ability to find vulnerabilities, exploit them and carry out cyberattacks, and all that with limited human intervention. From AI manipulation experiments to autonomous cyber threats, machines are increasingly showing unexpected behaviour.
Trending Videos
)
){kind=logo}
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)