Threat modeling in the age of AI - 3 steps to understanding AIML system attack surfaces

In 2024, data is no longer the “new oil”--it’s the new attack vector. How can security pros model these new threats, as AI becomes increasingly embedded in daily life? This session covers ways in which securing AI systems requires new paradigms of thought, including tips to proactively manage AI security risks, as well as practical ways to bridge knowledge silos, increase buy-in, and achieve tangible improvements in your AI security posture.

Full recap & slides: https://threatmodelingconnect.discour...

____________________________

Subscribe to our newsletter: https://www.threatmodelingconnect.com...
____________________________

Threat Modeling Connect is the world’s largest community of threat modeling practitioners. Our mission is to make threat modeling accessible to all through events, expert content, and free-to-use resources.

EVENTS
► Upcoming events: https://lu.ma/threatmodelingconnect
► Recordings: https://www.threatmodelingconnect.com...

RESOURCES
► Member newsletter: https://www.threatmodelingconnect.com...
► Community forum: https://threatmodelingconnect.discour...
► Guides: https://www.threatmodelingconnect.com...
► Blog: https://www.threatmodelingconnect.com...
► Threat model collections: https://www.threatmodelingconnect.com...


FOLLOW US
►LinkedIn:   / threatmodelingconnect  
►Twitter:   / tmconnecthq  


CHAPTERS
---------------------------------------------------
00:00 Announcements
03:23 Intro & the role of data in AI security
08:19 Three approaches to understanding AI threats
28:45 Three steps to mapping & securing the AIML system attack surface
40:09 Take home message