IntegrityShield designs assessments where AI-assisted answering fails
keviv9 · x · 2026-09-21
IntegrityShield, from ASU's CogniShield line of work, explores AI-resilient assessment design: instead of relying on detecting AI use after the fact, it investigates designing assessments where straightforward AI-assisted answering becomes less useful. A demo is available.
More from Apps
- First AI agent purchase via Muse and Link: live browser session, approval flows, one fix — jeff_weinstein · 2026-09-21
- Tesla's FSD Supervised gets approved in Czechia, rollout starting soon — bendomingueztv · 2026-09-21
- How to turn Claude into your first employee before you can afford one — evielync · 2026-09-21
- px0 IDE boots in <1ms with 20MB memory, winning users over from Zed — arpit_bhayani · 2026-09-21
- Wiring Muse to the iPhone Action Button: the fastest way into an agent yet — SeanOliver · 2026-09-21
- "Distribution is the bottleneck": dev says Muse is just OpenClaw for normies — 0xsachi · 2026-09-21