The Agentic Playground: How Threat Actors are Weaponizing Identity at Scale | SailPoint Navigate
Source: YouTube · SailPoint · published Oct 16, 2026 · 18:00
BLUF: The presentation analyzes a recent AI-augmented espionage attack by Midnight Blizzard to highlight critical gaps in identity governance for autonomous agents, advocating for dynamic runtime controls and comprehensive discovery to secure AI ecosystems 2:15.
Key Takeaways:
• A significant governance gap exists, with 48% of organizations deploying AI without redesigning governance controls, and 69% reporting limited control over AI autonomy 2:42.
• The featured attack utilized a device code flow via captive portals to register attacker devices, exploiting the lack of real-time authentication verification 5:30.
• The attack scale was achieved because only 15% of the operation required human intervention, while AI agents autonomously managed the remaining 85% of the infrastructure and timing 8:50.
• Successful defense requires moving beyond static policies to dynamic runtime controls that evaluate context, purpose, and behavioral drift of agentic identities 13:40.
• Organizations must first conduct comprehensive discovery of all human and non-human identities to establish a baseline before implementing risk mitigation and guardrails 15:50.
Closing Statement:
Securing the future of AI integration requires a fundamental shift from static identity management to dynamic, context-aware governance frameworks that can detect and mitigate risks posed by autonomous agentic behaviors.
Sources:
- 0:14 Introduction to the session on AI augmented espionage attacks.
- 2:15 Discussion on the growth of AI and the lack of governance controls in organizations.
- 3:14 Founding principles of identity security for the AI era.
- 5:30 Details on the device code flow attack vector used by Midnight Blizzard.
- 8:50 Explanation of the high level of autonomy in the attack execution.
- 13:40 The necessity of dynamic runtime policies over static ones.
- 15:50 Recommendation for comprehensive discovery of human and non-human identities.
Generate CPE Credits
Generate a professional CPE document from this video's transcript.
Estimated credit: 0.5 CPE hours
Estimate uses the video runtime (1 hour ≈ 1 CPE, rounded to the nearest 0.5, minimum 0.5, maximum 2.0). The final amount can be lower after review, never higher.
Topic: Identity & Access Management. Commonly maps to: Identity and Access Management (IAM), Security Architecture and Engineering. Exact CISSP domains are assigned during generation.
Free account. One generation at a time, with a daily limit.
CPEBuddy is independent and not affiliated with or endorsed by ISC2, ISACA, or any certification body. Exports are formatted for common CPE submissions; acceptance is at your certification body's discretion.
Transcript Preview
First 800 characters of the transcript
Good afternoon all. Thank you for joining us for this session on the agentic playground. We're going to dive into a live AI augmented espionage attack, an exploit that was was executed just a few months ago. Um, at the same time, we'll tie that back to identity and look at it from the lens that you all know well. Um, my name is Tim Carroll. Um, I'm a senior manager, uh, associate vice president in the identity security practice at Deote. Um, I've spent most of my tenure for the last decade primarily focused in Salepoint but also the broader uh, cyber security context as well. >> Hi everyone, my name is Aik Basak. I'm a similarly I'm a AVP in our engineering practice and focusing on identity. I have been doing identity for quite some time and focusing on NHI and privilege access. That's my …