AI fakers exposed in tech dev recruitment: postmortem.

Today's reference article comes from a different source than our usual one, as it is a digression published within The Pragmatic Engineer newsletter, run by Gergely Orosz, writer of “The Software Engineer's Guidebook.”

Artificial intelligence is making its way within the work sector with excessive ease, and not only as a valuable tool that can speed up tedious and repetitive processes, but as a potential candidate for positions open to real professionals. Recently, Polish startup Vidoc Security fell victim to a similar attempt, made by as many as two different candidates.

The entire recruitment process took place according to the general standards of the field, starting with a technical interview, followed by a coding testing phase and a final interview to confirm the candidate's availability. In the case at hand, the AI candidates mentioned proved to be perfectly capable of passing the first two stages, however, showing difficulties during the final interview, as the deepfake avatars used failed to stand up to direct confrontation with the interviewer.

The article explores the incident in greater depth, also providing a number of useful tips to avoid becoming a victim of these devious strategies.