❮ Back to all resources
Webinar

The New Supply Chain: Third-Party Risk in the Age of Apps, Agents, and AI

See how frontier AI finds and combines weak links across third-party integrations, and why security teams must inspect the access that remains after vendor approval.

Frontier AI models are reducing the expertise and research time required to find a way into an organization. They do not need to understand the business. They need to identify one weak link, then combine vulnerabilities, misconfigurations, and trusted connections into an attack path.

Those paths often begin with third-party integrations that already hold valid access. OAuth grants, service accounts, and app connections can gain permissions, reach additional systems, and remain active after their original owners move on. The vendor may still be trusted while the access behind it has changed.

In this 20-minute session, Sadia Haque and Chris Fuller examine why point-in-time vendor assessments miss these changes, the signals security teams need to detect them, and how to prioritize the integrations that pose the greatest risk.

Watch to learn:

  • How frontier AI finds and combines weaknesses across connected applications
  • Where permission growth, owner drift, and behavior changes create risk after vendor approval
  • How to map OAuth grants, service accounts, and cross-app connections to see what each integration can reach

Speakers

Sadia Haque, Obsidian Security

Chris Fuller, Obsidian Security