As AI becomes central to national security, alignment itself may introduce new risks. This week, Elisa Poteat is joined by Dr. Michael Vaiana, Research Director at AE Studio, to examine how AI alignment, model testing, and system integration shape both defensive and offensive security capabilities. Together, they explore red teaming, data poisoning, under-researched risks, and what policymakers need to understand about how AI systems behave when deployed at scale.

Dr. Michael Vaianais a Research Director at AE Studio

References:

The Dwarkesh Podcast: Ilya Sutskever – We're moving from the age of scaling to the age of research. Nov. 25 2025.