Gary McGrawAutonomy. where #MLsec escapes human control
By now we should ALL know this happens. Florida is behind...
Fable vs mythos...does benchmark result even mean anything?
Nope. #MLsec reality is more complicated than a benchmark
noplasticshower@Techaltar suleyman pretends recursive pollution does not exist #MLsec
@4Dgifts Ah yes. The #MLsec fun and games.
Did you read this yet? https://berryvilleiml.com/results/no-security-meter-ai.pdf
Darkreading takes on the #AI worm. Commentary by BIML (Thai is, me) #MLsec
https://www.darkreading.com/cyber-risk/adaptive-agentic-ai-worms-enterprise-cyber-threat
On Episode 157 of the Silver Bullet Security Podcast, BIML’s Gary McGraw hosts Tim Schulz. Tim talks about whitebox control and observability in machine learning systems (and especially transformer architectures), the limits of red teaming for securing AI, “neural surgery,” Agentic AI and the confused deputy problem, and the economics of network “smallification.” #AI #ML #MLsec
https://berryvilleiml.com/2026/06/01/silver-bullet-security-podcast-157-tim-schulz/
What exactly does BIML work on all day? Listen to this podcast and find out. #DataScience #data #MLsec
Socks, Crocs, and AI Security | Podcast Episode on RSS.com
As a founder of the Berryville Institute of Machine Learning, Gary McGraw has been researching AI security since before most people knew what machine learning was. He's identified 78 risks across ML systems and was sounding the alarm on recursive pollution and model collapse long before those terms went mainstream. He joins Sid and Lee to break down what practitioners need to understand about the systems they're implementing, why 23 of those risks live in a black box controlled entirely by the foundation model vendors, and what good governance looks like when you can't see inside the thing you're governing.