Part of me suspects that this is a major reason for the push on RTO. If you hire people remote who used any one of dozens of mechanisms to cheat.. it can be hard to correct. You could have a candidate who does great on your interview then fails to produce a single PR in 3 months on the job.

Speculation aside, this means you have a bad interview process.

If your interview isn't screening for the work you want somebody to do then it's not a good metric for finding successful employees.

I think the fraud is asymmetric. It's relatively easy for me to hire someone to perform the work the employers wants for X hours one time (i.e. interview), but relatively difficult for me to subcontract it reliably for 40 hours every week.

This asymmetry exists regardless of how representative of real work the interview process is, short of a full-time engagement.

If he fails to deliver on the job, put them on a PIP and get rid of them.

If he actually delivers good work, who cares if he cheats?

The loss is three months of salary and the cost of hiring a replacement. Taken en masse, especially for smaller firms, quite a debilitating sum.

Is he actually doing the work? Is just copying code from OSS projects and does not actually understand it? Not only will this cause legal liability, but when the team is under stress will he just disappear or be able to solve an issue?

He already lied once. Why not lie more?

Lying is already rampant on both sides of the hiring process, and expected. The high stakes and algorithmic nature of (tech) hiring is the real problem - the rewards are great and the exploits are obvious.