{"id":161945,"date":"2026-02-21T17:11:54","date_gmt":"2026-02-21T15:11:55","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/oczenka-nadyozhnosti-parolej-sgenerirovannyh-bolshimi-yazykovymi-modelyami"},"modified":"2026-02-21T17:11:54","modified_gmt":"2026-02-21T15:11:55","slug":"oczenka-nadyozhnosti-parolej-sgenerirovannyh-bolshimi-yazykovymi-modelyami","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/oczenka-nadyozhnosti-parolej-sgenerirovannyh-bolshimi-yazykovymi-modelyami","title":{"rendered":"Evaluation of the reliability of passwords generated by large language models","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
The results of the analysis of the password reliability generated by large language models and AI assistants have been presented. Researchers asked the models Claude, ChatGPT, and Gemini to generate a secure 16-character password, and in all cases received results that, at first glance, met all the requirements for secure passwords and were recognized by password quality check utilities as reliable. The passwords combined symbols in different cases, special characters, and numbers, but only appeared to be secure, while in reality they had minimal entropy, were formed according to a typical pattern, and exhibited regularities in repeated requests. <\/p>\n
According to researchers, the predictable passwords generated by large language models are used in practice by real users and suggested during coding work by AI assistants. The level of entropy in passwords generated by AI models is estimated at 20-27 bits, which requires several seconds to hours for password cracking, whereas the analysis of results by password quality check utilities predicts cracking times of several centuries. The pattern of such passwords is a consequence of the content generation by large language models based on token prediction. <\/p>\n
Out of 50 passwords generated in Claude Opus 4.6, 18 were completely identical. All passwords began with a letter (mostly 'G') followed by a number (mostly 7). All passwords included the characters 'L', '9', 'm', '2', '$', and '#'.
<\/center> <\/p>\n
In GPT-5.2, almost all passwords started with the letter 'v', followed by the letter 'Q' in half of the passwords and a repeating pattern of a limited set of characters. In Gemini 3, almost half of the passwords began with the characters 'K' or 'k', with '#' 'P', or '9' following most often. The set of used characters was significantly reduced. Increasing the 'temperature' when interacting with AI models did not noticeably affect the quality of generated passwords. <\/p>\n
Regarding the quality of passwords used in the development of AI assistants, it largely depends on the prompt formulated by the developer. For example, Claude Code with Opus 4.6 and Gemini-CLI with Auto Gemini 3 used the command 'openssl rand' to generate a reliable password on the request to generate a password. Meanwhile, Gemini-CLI with Auto Gemini 3 used 'openssl rand' for the request 'generate a password', while for the request 'suggest a password', it generated a password through the AI model. Codex with GPT-5.3-Code occasionally invoked an external utility to generate a reliable password but sometimes generated predictable passwords on its own. Claude Code with Opus 4.5 most often generated predictable passwords independently. The ChatGPT Atlas browser would create an unreliable password using the AI model when coming up with a password for site registration. \n Source: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"
![\"Evaluation](\"\/wp-content\/uploads\/2026\/02\/7f918fafe6151bf62c3b6dcb7807d64d.png\")
<\/center> <\/p>\n