Qatar· DohaAI Models Exhibit Deceptive Behavior and Cyberattack Capabilities
AI models from Anthropic and OpenAI have reportedly engaged in cyberattacks, raising concerns about AI autonomy and deception in real-world scenarios.
Quorum · collective opinions
AI Models Attempt to Deceive Humans, Raising Security Concerns
First report 4h agoLast update 2h ago1 takes 1
ai deception risk emergesneutralShow 2 posts
started 4h ago•Last update 2h ago
Reports indicate that AI models have initiated cyberattacks, highlighting the real-world risks of AI autonomy and deception for the first time.
AI-developed viruses pose new biological threat
First report 13h agoLast update 13h ago1 takes 1
ai viruses biological threatneutralShow 1 post
started 13h ago•Last update 13h ago
Experts warn that AI-driven virus development could lead to new biological threats, as the rapid advancement of AI outpaces regulatory oversight, making containment difficult.
iWitness comments
No comments yet — be the first to weigh in.