WEBINAR ON DEMAND
Duration: 1 hour

AI-based applications are reshaping business strategies, but despite the benefits, they bring significant cybersecurity risks. Pentesting reveals the unique AI vulnerabilities that traditional security measures frequently miss. According to recent insights from thousands of tests via Cobalt’s Offensive Security Platform, LLM pentests are 2.5 times more likely to reveal high-risk findings than other types of pentests, and only 21% of serious findings from these tests will get resolved.

Join cybersecurity experts Jason Lamar (SVP of Product, Cobalt) and Wade Baker (Co-Founder, Cyentia Institute) as they discuss findings from Cobalt’s State of Pentesting 2025 Report. They’ll share why organizations must urgently adapt their security strategies to the AI-driven landscape. This webinar covers:

• Unique vulnerabilities posed by AI-based applications
• The most common LLM security issues identified by pentesting
• Differences between AI security and AI safety—and why both matter
• How pentesting reveals critical gaps in organizational security postures
• Best practices for addressing and mitigating AI-specific vulnerabilities

SPEAKERS

Jason Lamar
​SVP, Product
Cobalt
 

Dr. Wade Baker
Co-Founder
Cyentia Institute