Detect candidates that cheat with AI / ChatGPT

Our proprietary detection mechanism

Coderbyte is committed to ensuring that cheating is both prevented and detected in order for organizations to make data-driven and unbiased hiring decisions. Many organizations are concerned about AI being yet another opportunity for candidates to cheat on assessments. 

Note: Despite the overwhelming excitement and extraordinary claims about AI capabilities on Twitter and other often reliable news sources, fewer than 5% of challenges on Coderbyte can currently be solved by ChatGPT without making any modifications to the code produced. Nevertheless, out of an abundance of caution, we've invested heavily to detect AI usage using proprietary methods.

On your assessment settings page, you will see an option for AI / ChatGPT detection automatically turned on.

Similar to plagiarism detection, Coderbyte will automatically track suspicious behavior and scan your candidates' solutions for a match with AI / GPT solutions for each challenge.

If a candidate used AI to cheat on an assessment, you will see it at the top of the candidate report:

You will also see within the relevant challenges that AI was used:

If you click Expand you will see the cheating that was detected and a link to the corresponding ChatGPT output that the candidate copied: