We Might Need AI to Defeat AI

AI presents many potential positives for both business and personal lives, but it also comes with many risks as well. We’ve seen how AI can be used to improve the abilities of bad actors, both through the development of malware as well as the ability to craft convincing phishing campaigns.

As AI cyberattacks become more prevalent, and their capabilities become more impressive, it’s possible the best approach to fighting AI will be hiring AI in the first place.

That’s what Jericho Security aims to do. As reported by VentureBeat, the cybersecurity company raised $3 million in pre-seed funding based on their pitch to develop AI systems that will combat cyberattacks powered by AI. The idea is to simulate situations between the two AI factions to understand where the weak points are.

Jericho Security plans to do this by creating an “AI red team” and “AI blue team:” The red team will play the side of the aggressor, developing cyberattacks from AI, while the blue team will figure out how to prevent attacks. When brought together, the company hopes to create a “feedback loop” that will unearth new angles that weren’t seen before.

Generative AI attacks are already on the loose. Jericho Security cofounder and CEO, Sage Wohns, highlighted cases involving fake invoices from both Costco and Best Buy, phony bank account updates, and illegitimate VC offers for start up companies like his.