Business Problems We Solve: AI Attack Vulnerability Reviews

In their wonderful recent book Not With A Bug, But With A Sticker: Attacks on Machine Learning Systems and What To Do About Them, Siva Kumar and Anderson note that "major AI systems remain vulnerable to the exploits of bad actors of all stripes." 

Including yours.

Siva Kumar and Anderson note three fatal flaws of managerial thinking when it comes to preparing for cybersecurity attacks by artificial intelligence (AI) agents. First, they note, cybersecurity managers have a "romanticized view" of attackers (e.g., hoodie-wearing nerdy teens living in mommy's basement). The authors recommend that cybersecurity managers broaden their view of who their adversaries could be.

Separately we note that any and all organizations are potential AI-based cybersecurity targets -- something alluded to by Siva Kumar and Anderson.

Second, the authors state that most people are AI-illiterate. Proof? Oh sure, you can ask an AI to create a deviled eggs recipe, but do you know how to evaluate the response? Probably not. And here's a Pythia Cyber pro tip: the only thing the AI platform guarantees you is that it could be wrong. With eggs being so expensive, do you want to take a chance on a bad recipe? Now multiply that by the cost and scope of your chatbot-derived risk management plan.

Third, the authors claim that organizations are so stove-piped in organization that they are not well-equipped to address these problems systemically. It falls upon IT...who might not be ready or be able to explain the problem...

Cybersecurity risk management systems are at risk for AI-based attacks. We solve this problem through assessing your vulnerabilities to AI-based attacks. 

While no system anywhere is always secure from AI-based attacks, at the same time nearly all systems need the type of review and planning we offer. Ask us how we can work with you to help you review and mitigate your AI attack vulnerabilities.