There are already prototypes
Posted: Tue Feb 11, 2025 4:27 am
A recent Forrester study found that 88% of security professionals expect AI-powered attacks to become mainstream in an era of what is already a huge shift in the nature of cyberattacks. Almost half of those surveyed believe it will happen within the next year and that it is only a matter of time. Open-source projects focused on AI research, tools that can be used to improve every phase of the attack lifecycle, already exist and will undoubtedly soon be among the paid hacking services available on the Dark Web.
of “offensive” AI that, in a matter of seconds, autonomously select the most profitable targets based on information about organizations gleaned from social media. The AI then creates context-sensitive cayman islands whatsapp data emails, selects a deceptive sender name, and sends out messages in an attempt to entice victims to click a malicious link or open an app that will expand their access to the organization’s network. These tools have been tested against “defensive” AI, simulating what we expect to see in the real world soon: AI versus AI in an algorithmic war.
Using this research and data, defensive AI gains additional capabilities. Using unsupervised machine learning, it has a holistic understanding of every user and device on the network it is protecting, and uses this increasingly complete understanding to spot small anomalies that could be signs of an incipient attack. With this bird’s eye view of digital business, defensive AI will spot offensive AI as soon as it starts manipulating data.
of “offensive” AI that, in a matter of seconds, autonomously select the most profitable targets based on information about organizations gleaned from social media. The AI then creates context-sensitive cayman islands whatsapp data emails, selects a deceptive sender name, and sends out messages in an attempt to entice victims to click a malicious link or open an app that will expand their access to the organization’s network. These tools have been tested against “defensive” AI, simulating what we expect to see in the real world soon: AI versus AI in an algorithmic war.
Using this research and data, defensive AI gains additional capabilities. Using unsupervised machine learning, it has a holistic understanding of every user and device on the network it is protecting, and uses this increasingly complete understanding to spot small anomalies that could be signs of an incipient attack. With this bird’s eye view of digital business, defensive AI will spot offensive AI as soon as it starts manipulating data.