The Future of Threat Detection and Response
summary
Argues that adversaries now target routers, switches and firewalls to “stay in” and turn systems off, that AI models never forget and are non‑deterministic so prompt attacks can extract secrets, and that telemetry grows by 1000X so one monster lake is impractical. Proposes distributed storage + federation, analytics pushed to the edge, and distributed enforcement via eBPF and a smart switch DPU/GPU model. Notes Log4j still exploited years later and shows early demos (Splunk/Cisco booths).
AI-written summary · about
chapters (8)
- 0:00Intro
- 1:52Artificial Intelligence
- 3:40Traditional Security Architecture
- 5:10Distributed Security Architecture
- 8:07Data Analytics Architecture
- 9:13Distributed Policy Enforcement
- 13:10Exploit Acceleration
- 14:51Where is AI going
related talks
45mSecuring The Future
55mKeynote - The Future of Cybersecurity Needs You, Here is Why.
58mKeynote: The Future of Incident Response
45mDefenders Struggle as the Adversary Continues to Break Away
24mThreat Response Needs New Thinking. Don't Ignore This Key Resource.
45mFast, Ever-Evolving Defenders: The Resilience Revolution
48mThe Future of Security isn't Preventing Attacks
05mSponsored Keynote: Cloud Security’s Hidden Force: Threat Detection