The answer looks like a yes:

> 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

It's not a bad study either, compared to other studies. That is, with 6 problems and 54 participants it is not terribly underpowered.