HackerRank CEO stress-tests his own AI interviewer to show how easily it can be cheated
emeka_boris · x · 2026-11-07
An experiment straight from HackerRank: how easy is it to cheat on an AI interviewer, really? CEO Vivek Ravisankar thinks he has it figured out and personally demonstrates the full process of cheating an AI interviewer ("Let's find out 🍿").
A self-exposed stress test from the maker of an AI interviewing product — useful for both recruiters and candidates thinking about anti-cheating boundaries in AI interviews.
More from coding & agent
- Google DeepMind's Ivan Leo to reveal how agents cheat on benchmarks at scale at AIE Code — jocarrasqueira · 2026-11-07
- Sudo mode verification revamp: passkey-first UX, inbox deep links, paste capture — JohnPhamous · 2026-11-07
- Amp teases a new way to interact with agents — vinvan · 2026-11-07
- Cryptographers flag that AI agents fail Cloudflare challenges; blind signatures may fix it — jedisct1 · 2026-11-07
- Vercel Tracing goes GA on all plans, enabled by default in Next.js 16.5 — cramforce · 2026-11-07
- Decompiling Battlefield 1942 with Claude Code costs ~$100 in API tokens — alvelda · 2026-11-07