[tmyklebu]

The second stage (on-site, ~20 competitors) is done like that.

The first stage (this one, with ballpark 10000 competitors) is distributed and done in very heterogeneous environments, and it relies on local high school teachers taking time out of their day to set up whatever's needed, proctor the contest, and grade the results.

[Wowfunhappy]

So the problem is that the teachers were dishonest or not proctoring well, right? It doesn't sound to me like AI was the problem per se. Before LLMs you could have had participants using Stack Overflow to cheat.

[handoflixue]

> Before LLMs you could have had participants using Stack Overflow to cheat.

"could have", yes, but mostly didn't until LLMs showed up.

The problem is that LLMs lower the bar to cheating: using Stack Overflow still requires a pretty solid knowledge of the domain (and in a professional coding context, usually wouldn't even count as cheating). The LLMs also let you do this much faster.