Everyone fears AI writing malware. The real risk is more boring: it ships an unauthenticated route to your production DB and calls it done. Looks finished. Runs fine. Wide open. What's the 1st thing you check in AI-generated code?
@hthieblot Building a service to scan vibe coded apps/sites for vulnerabilities. Too many people are using the tool that built their app to then scan it for vulnerabilities.
We have a 5 stage process that is very complex.
@hthieblot We've built a website access log system. We ingest the logs, analyze them, report the malicious IPs to the responsible org and track everything. We have also built our own LLM that positively identifies website malware. It's better, faster, stronger than those using signatures.
I kept hearing about how "everyone" is using Fail2ban and/or Imunify360 to block attacks on xmlrpc.php and wp-login.php on their WordPress sites.
Here's the truth.
https://t.co/3DJxhYA7qW