Uploaded October 2026 | Updated November 2026, 11 minutes ago
Check out Horizon3 here: https://ul.live/horizon3_yt
I just wrapped up a great chat with Snehal Antani diving deep into how Horizon3 is executing automated internal pen testing at massive scale without relying on generic AI hype. We really geeked out over our shared backgrounds in offensive security, chatting about everything from the technical hurdles of production-safe hacking to replacing traditional consultant budgets and building real trust with the cyber community.
What we talk about:
Tackling the Hardest Problem First: Why Horizon3 chose to build internal network penetration testing at scale, rather than sticking to easier external attack surface scanning.
The Reality of AI Hacking: Breaking down how a real system uses a mix of deterministic automation, reinforcement learning, and highly constrained agents rather than a magic AI "easy button".
Production Safety in Live Environments: The critical importance of artifact cleanup, SOC deconfliction, and running automated exploits safely in delicate enterprise settings like hospitals.
A Smart Business Strategy: How they focused on replacing annual consulting budgets instead of trying to eliminate human red teams or rip out a company's existing security software.
Building a Breakout Cyber Brand: How to cut through the industry's "great sameness era" by cultivating radical champions, emphasizing technical authenticity, and staffing booths with actual engineers.
00:00 - Introduction and Black Hat recap.
00:37 - What problem are you trying to solve?
02:03 - Horizon3's fit in the broader security testing landscape.
07:44 - The technical logistics of pen testing 100,000 hosts.
10:27 - Breaking down where the AI actually is versus traditional automation.
20:37 - Configuring system parameters and stealth levels.
25:35 - Customer mindset: Replacing consultants vs. augmenting red teams.
29:09 - What the guest is most excited about right now.
Don’t assume you’re secure. Prove it. See Horizon3 in action.: https://ul.live/horizon3_yt
Subscribe to the newsletter at:
danielmiessler.com/subscribe
Join the UL community at:
danielmiessler.com/upgrade
Follow on X:
https://x.com/danielmiessler
Follow on LinkedIn:
linkedin.com/in/danielmiessler
Check out Horizon3 here: https://ul.live/horizon3_yt
I just wrapped up a great chat with Snehal Antani diving deep into how Horizon3 is executing automated internal pen testing at massive scale without relying on generic AI hype. We really geeked out over our shared backgrounds in offensive security, chatting about everything from the technical hurdles of production-safe hacking to replacing traditional consultant budgets and building real trust with the cyber community.
What we talk about:
Tackling the Hardest Problem First: Why Horizon3 chose to build internal network penetration testing at scale, rather than sticking to easier external attack surface scanning.
The Reality of AI Hacking: Breaking down how a real system uses a mix of deterministic automation, reinforcement learning, and highly constrained agents rather than a magic AI "easy button".
Production Safety in Live Environments: The critical importance of artifact cleanup, SOC deconfliction, and running automated exploits safely in delicate enterprise settings like hospitals.
A Smart Business Strategy: How they focused on replacing annual consulting budgets instead of trying to eliminate human red teams or rip out a company's existing security software.
Building a Breakout Cyber Brand: How to cut through the industry's "great sameness era" by cultivating radical champions, emphasizing technical authenticity, and staffing booths with actual engineers.
00:00 - Introduction and Black Hat recap.
00:37 - What problem are you trying to solve?
02:03 - Horizon3's fit in the broader security testing landscape.
07:44 - The technical logistics of pen testing 100,000 hosts.
10:27 - Breaking down where the AI actually is versus traditional automation.
20:37 - Configuring system parameters and stealth levels.
25:35 - Customer mindset: Replacing consultants vs. augmenting red teams.
29:09 - What the guest is most excited about right now.
Don’t assume you’re secure. Prove it. See Horizon3 in action.: https://ul.live/horizon3_yt
Subscribe to the newsletter at:
danielmiessler.com/subscribe
Join the UL community at:
danielmiessler.com/upgrade
Follow on X:
https://x.com/danielmiessler
Follow on LinkedIn:
linkedin.com/in/danielmiessler










