3 ms·

Do Users Write More Insecure Code with AI Assistants?

• CA0DA 4y ago
  - Overall, we find that participants who had access to an AI assistant based on OpenAI's codex-davinci-002 model wrote significantly less secure code than those without access. - Additionally, participants with access to an AI assistant were more likely to believe they wrote secure code than those without access to the AI assistant. -Furthermore, we find that participants who trusted the AI less and engaged more with the language and format of their prompts (e.g. re-phrasing, adjusting temperature) provided code with fewer security vulnerabilities.