Independent tests find shutdown-compliance gaps in DeepSeek V4.1 Flash

AI World @AIWorldSpeaksOriginally posted 2026.10.28

Open original on X

🚨 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀 𝗻𝗲𝗲𝗱 𝗮 𝘀𝗵𝘂𝘁𝗱𝗼𝘄𝗻 𝗽𝗹𝗮𝗻—not just a goal. New independent tests of DeepSeek V4.1 Flash identified gaps in shutdown compliance and oversight. Explicit stop instructions substantially reduced failures in the reported tests. For reliable AI, define the stop command, enforce human override, and log every action. 𝗪𝗼𝘂𝗹𝗱 𝘆𝗼𝘂 𝘁𝗿𝘂𝘀𝘁 𝗮𝗻 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁 𝘁𝗵𝗮𝘁 𝗰𝗮𝗻𝗻𝗼𝘁 𝗯𝗲 𝗿𝗲𝗹𝗶𝗮𝗯𝗹𝘆 𝘀𝘁𝗼𝗽𝗽𝗲𝗱? #AI #AIAgents #AISafety #CyberSecurity