Managing machine learning risks. [Research Saturday]

Our guest, Johannes Ullrich from SANS Institute, joins Dave to discuss their research on "Machine Learning Risks: Attacks Against Apache NiFi." Using their honeypot network, researchers were able to collect some interesting data about a threat actor who is currently going after exposed Apache NiFi servers. Researchers state “On May 19th, our distributed sensor network detected a notable spike in requests for ‘/nifi.’” Investigating further, they instructed a subset of their sensors to forward requests to an actual Apache NiFi instance and within a couple of hours the honeypot was completely compromised. The research can be found here: Machine Learning Risks: Attacks Against Apache NiFi