A Decision Model Breaks Like Any Other Language Model: A First Look at Jev

Summary

Check Point researchers tested Jev, a new AI model that returns typed decisions instead of text and is built for software to call. Working from outside a realistic application, they changed its verdict, downgrading risk to low and advising investment on a document flagging every warning sign. They report about 50 cents per successful break.

Related items

• InfoImproving Fuzzy Robustness in LLM Safety with Automated Red TeamingSimilar attack · OpenAI Alignment Research Blog
• InfoBait: Mitigating Harmful Agentic Reinforcement Learning via Probing Reward FunctionSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• InfoImage watermarking robustness under MLLM-guided semantic editing: Benchmark, defense, and diagnosisSimilar attack · Elsevier Security Journals
• LowYour Robot's Safety Functions Already Work. What If the Input Lies?Similar attack · VicOne Blog
• InfoSpeech Tokenizers are Vulnerable: Transferable Semantic Attack and Robust TokenizerSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)