AI-Driven Cyber-Attacks

This talk intents to provide thought leadership on how AI & ML is currently being used for defensive purposes and will soon be used for offensive purposes. It covers both sides, defence and offence, and provides examples of how AI either is currently used to augment defensive measures or how AI will be used to augment cyber offence.

On the defensive side we have seen an explosion in the vendor landscape using narrow AI to varying levels of success. What is undeniable is that there is improvements in blue teaming & cyber defence thanks to AI. Max will give a handful of practical examples of where AI has caught and stopped extremely sophisticated cyber attackers. AI in defence is already a reality.

On the other hand, cyber attacks are becoming increasingly fast and sophisticated. WannaCry & NotPetya denominated a first paradigm shift - from low & slow attacks to fast-moving, spreading & destructive malware.
Relying on signature-based / hard-threshold-based & maintaining the mentality of 'keeping bad out' instead of 'assuming breach' is not adequate in today's threat landscape any more.

Cyber criminals adopt whatever method yields the biggest profits. AI will become another tool in the repertoire of adversaries allowing them to:
* Improve scalability & profitability
* Stay undetected for longer & make attribution more difficult
* Conduct cyber attacks a human attacker could not do

This talk will further provide insights on how adversaries could leverage existing AI research & Open-Source projects to augment their attacks along the attack lifecycle. This will help people understand that the cyber security & threat landscape is rapidly changing and that organisations have to start looking into novel ways of conducting cyber security to be prepared when the next paradigm shift hits our industry - AI-driven cyber attacks.

Numerous malware variants are being created daily. To adjust to this evolution, machine learning tools are being utilized by security companies to detect the novel threats and new attack vectors. Same for the threat hunting, where the ML helps in proactively and iteratively parsing through networks detecting the advance threats. Important question is where we want to apply these advanced techniques. The technology should be applied in a smart way to tackle specific problems. In this panel we will discuss the current state of AI in cybersecurity and what the future will hold.