Last Night A DJ Erased My Drive: A Journey Through Cybersecurity, Chaos, And Control

Mathew Caplan builds the talk around an analogy between music and cyber security, arguing that both start with a trigger, unfold over time and depend on timing and rhythm. He compares a security stack to a recording studio, where threats are the signal, alerts and false positives are the noise, and the architecture is the mix. Early sections survey how music has been used as propaganda, as a tool of coercion and as a form of resistance, from coded songs to bootleg records cut from discarded X-ray film.

The middle of the talk turns to modern attacks. Caplan covers AI-generated tracks and streaming fraud as an example of abuse at scale, then walks through the 2020 SolarWinds supply chain compromise and the 2024 Live Nation data breach, where stolen credentials and missing multi-factor authentication let attackers log in as trusted users in a third-party environment. The common thread is that identity-driven intrusions look like routine activity, move slowly and stay below alert thresholds for long periods.

For practitioners, the closing recommendations include zero trust with continuous checks on identity and context, reducing noise through least privilege and segmentation, and managing third-party risk with priority on critical suppliers. He advises detecting attack sequences rather than isolated events, scoring alerts by risk, correlating signals across domains, and running purple teaming against realistic scenarios so that detections are actually tested. The main message is that simpler architecture and pattern-based detection matter more than chasing every signal.