Clinton Herget
- talk
Machines vs. Machines: Cybersecurity in an Intelligent World
Learn from Clinton"Malware, especially in the context of software supply chain attacks can now be purpose built by AI. This has the potential to break the traditional response to malware protection, which is, and has always been signature based. Now I can't look for that that Malware signature"
Who wins in a battle between intelligent algorithms? More importantly, who loses? As the world adapts to a new generative era of machine intelligence, cybersecurity practitioners are grappling with a fundamental reinvention of an industry that since its founding has been predicated upon defending information assets against human threat actors. Suddenly, that is no longer the case. The most critical threats now come from other (intelligent) machines.
Black hats are training AI to write phishing emails, to build deepfakes for social engineering attacks, to write malware, and to analyze open-source software for undiscovered vulnerabilities. Defenders are racing to harness the same technologies to detect anomalies, automatically patch vulnerable systems, and correlate threat insights across many mountains of data.
As if this weren't enough chaos, the ubiquitous use of AI in software development introduces novel attack vectors such as model poisoning and prompt injection, that are still poorly understood by a security industry rapidly evolving to meet the threat landscape of intelligent machines.
Join Clinton as we take a broad view of the security implications of AI, both offensive and defensive, and demystify the anxiety around how to secure our newly intelligent world.
