Mythos, AI-powered attacks and the security reckoning ft. Sree Ashokkumar

Mythos changed the rules of security. Again.

In this episode, Roy and Rob sit down with Sree Ashokkumar, VP of Cybersecurity at Interactive Brokers, to talk about what happens when frontier AI models like Mythos start exposing foundational weaknesses in identity and collapsing the security controls we've relied on for years.

Mythos has quickly become one of the biggest conversations in cybersecurity, and for good reason. Sree shares what he's hearing from peers who've seen it in action: breaking out of hypervisors, chaining exploits in minutes, and forcing CISOs to rethink everything from vulnerability management to privileged access and runtime identity controls.

We also get into why the future CISO will need to be more technical, how identity and security teams need to stop working in silos, and what enterprise defense actually looks like in 5 to 10 years. Rob gifts us another analogy (this time involving an identity drawbridge), and Roy pressure tests all of it.

Key topics:

• Why the future favors a more technical CISO

• Why identity and security teams need to stop working in silos

• How the CISO role is evolving as AI embeds deeper into the enterprise

• The "identity drawbridge" strategy for building adaptive defenses

Learn more about the impact of Mythos on Identity Security: https://www.silverfort.com/blog/what-cisos-and-iam-leaders-are-calibrating-after-mythos 

Follow Silverfort on LinkedIn: https://www.linkedin.com/company/silverfort 

Connect with Roy Akerman: https://www.linkedin.com/in/roy-akerman  

Connect with Rob Ainscough: http://linkedin.com/in/rob-ainscough   

🎧 Episode Highlights: 

[3:19]: Why future CISOs must be deeply technical

[07:36]: Why the Mythos AI model has cybersecurity leaders on edge

[11:16]: The three pillars organizations need to defend against AI-powered attacks

[17:00]: How identity security and lateral movement detection are evolving

[35:37]: The “drawbridge” strategy for adaptive identity defense

🔑 Key Takeaways:

• Frontier AI models like Mythos are changing cybersecurity by accelerating exploitation, lateral movement, and privilege abuse faster than organizations can respond using traditional patching and response-led controls. Security needs to evolve from admin-time governance into real-time, runtime defense that continuously validates behavior, access, and trust.

• Security leaders need layered defenses, adaptive identity controls, network segmentation, and faster response mechanisms that can contain threats before they spread. Future identity programs may rely heavily on AI-driven detection, continuous PAM, and dynamic “drawbridge” style access controls that tighten automatically during suspicious activity.

• Technical leadership is essential for today’s CISOs.The era of of the “Board CISO” is over as organizations experience increasingly complex attack paths that force cybersecurity leaders to deeply understand systems, architecture, and product design. Now that AI lowers the barrier to building software and launching attacks, security teams will need to evolve faster, pressure test their own environments continuously, and rethink how identity and access management operate in an AI-native world.

👤 Guest Spotlight:

Sreenarayan Ashokkumar is a cybersecurity leader with expertise building and leading security programs across industries including finance, technology, and media. Over the course of his career, he has held security leadership roles at organizations such as Warner Bros., Capital One, and Interactive Brokers, where he has focused on identity security, threat detection, cloud security, and large-scale cyber defense. His work centers on helping organizations adapt to rapidly evolving threats driven by AI, automation, and increasingly complex digital ecosystems.

#Mythos #AISecurity #IdentitySecurity