Greetings,
We just got back from Defcon a couple of weeks ago where we gave a talk, "Attacks on GenAI data and using vector encryption to stop them." This one was a joint presentation with me and our CTO, Bob Wall, and we went a lot deeper on the encryption we use to handle AI vector embeddings.
Unfortunately, the video of the Defcon talk is currently only available to those willing to pay, but when it finally hits Youtube, we'll share the link.
If you haven't ever been to Defcon, you should really add it to your bucket list. They don't call it "hacker summer camp" for nothing. If you can brave the Las Vegas heat in August, you'll find a conference unlike any other. It is far more experiential with far more content and hands on learning than any conference I've ever attended.
We've been to three conferences in the last couple of months and have had tremendous response to our talks. Overwhelming feedback has been that no one is covering the broader AI security problems in depth and that they'd like more information to share with colleagues. So here it is! A bigger, better white paper on the Security of AI systems (now in technicolor!). (At the moment, you'll need to grab it from LinkedIn.)
Finally, if you're in the Bay Area, we'll be there the week of the 23rd for the OpenSearch Conference. Let us know if you'd like to meet up.