Frontier AI Is Killing Open CTF

A eulogy-cum-manifesto from the security community: frontier models' problem-solving has drained open-registration CTF competitions of meaning, and veteran players are bowing out.

How the Competition Got Wrecked

CTF (capture the flag) was once the security world's training ground and social scene, its charm being the contest between people and puzzles. Now frontier models' solving speed on routine problem types far exceeds humans', and in open-registration competitions, teams running AI scores crush people doing the problems in earnest, leaving the leaderboard without value. The author's anger isn't about the technology but the asymmetry: banning AI can't be verified, and allowing AI turns the competition into a contest of whose API budget is bigger—neither of which is CTF's original intent.

The Loss Beyond Competition

The deeper worry in the comments is the loss of the educational function: a generation of security engineers was "ground" into shape in CTFs, and the intuition accumulated in the solving process can't be gained by looking at answers. The same script has played out in Go, competitive programming, and online judge systems, and the response paths are similar too: shift to offline, real-name, controlled-environment competitions, and remake open events into human-AI collaboration formats. CTF won't die, but that era of winning respect through anonymous IDs and pure craft is indeed drawing to a close.

via: Hacker News