Critical Paperclip bugs expose AI agent trust failures

Security researchers are warning against trust assumptions in AI security with newly detailed flaws affecting the open-source AI agent platform Paperclip tha...