14What scans and admission policies would you require before an open-weights model runs in production?▼mediumNewHugging FaceDatabricksMeta◆ premiumOpen weights arrive with code, configs, licences and unknown training. The scored answer turns due diligence into a repeatable gate that production enforces automatically.Open full answer →
20How do you stop employees leaking company data into public AI tools?▼easyNewJPMorgan ChaseMicrosoftPalo Alto Networks◆ premiumBanning tools people find useful mostly moves the leak somewhere you cannot see. The scored answer offers an approved route, then adds policy, technical controls and visibility around it.Open full answer →
24Design abuse policy, reporting and enforcement for a consumer generative AI feature.▼mediumNewMetaGoogleOpenAI◆ premiumA trust and safety design question. The scored answer connects policy, detection, review, enforcement and appeals into one loop, and sizes it to the volume and the harm.Open full answer →
22Design an automated red-teaming agent that finds new failure classes in a target system from its policy.▼expertNew◆ premiumDesign a bounded assessment workflow from approved requirements to verified findings. Separate genuine mechanism discovery from new wording, scorer errors and unresolved effects.Open full answer →
09How do you govern shadow AI?▼easyNewJPMorgan ChaseMicrosoftSnowflake○ sign inBanning AI tools moves usage to personal accounts. The answer that works gives people a sanctioned path that is easier than the shadow one.Open full answer →