Uploaded August 2026 | Updated November 2026, 2 hours ago
Varun Badhwar told me 60% of AI generated code is insecure by default. So I asked why we should trust models to catch security issues when models are the ones causing them. His answer: on their benchmarks it's actually 70 to 80%, even from the most sophisticated models. Full episode: hanselminutes.com/1052 #shorts
Varun Badhwar told me 60% of AI generated code is insecure by default. So I asked why we should trust models to catch security issues when models are the ones causing them. His answer: on their benchmarks it's actually 70 to 80%, even from the most sophisticated models. Full episode: hanselminutes.com/1052 #shorts










