How is agentic AI changing development and security?
@fsmontenegro joins ๐๐ ๐ก๐ฎ๐๐ถ๐๐ฒ๐๐ฒ๐ to explore that shift through the lens of modern cybersecurity, attack surface challenges and evolving security architectures.
๐ March 23 at RSAC
๐ Included with RSAC badge
๐ buff.ly/AAnFn1e
#AINativeDev #RSAC #AgenticAI #Cybersecurity
@fsmontenegro joins ๐๐ ๐ก๐ฎ๐๐ถ๐๐ฒ๐๐ฒ๐ to explore that shift through the lens of modern cybersecurity, attack surface challenges and evolving security architectures.
๐ March 23 at RSAC
๐ Included with RSAC badge
๐ buff.ly/AAnFn1e
#AINativeDev #RSAC #AgenticAI #Cybersecurity
Mar 21, 2026 ยท 12:28 AM UTC
1
4
6
229
@TechstrongGroupc@fsmontenegro the one most agentic AI builders are ignoring completely. When you give agents SSH access, browser control, and permission to execute commands across multiple machines, you've created a distributed system with enormous privilege.
We run a 3-node Claude mesh and the security layer was an afterthought until our cybersecurity partner looked at it and said "you built a botnet that happens to be helpful." Credential vaulting per machine, permission boundaries per node, heartbeat monitoring for unauthorized access. It's not optional.
Most teams building agentic systems right now are focused on capability. Security should be the first layer, not the last. Excited to see this getting stage time at RSAC.
We run a 3-node Claude mesh and the security layer was an afterthought until our cybersecurity partner looked at it and said "you built a botnet that happens to be helpful." Credential vaulting per machine, permission boundaries per node, heartbeat monitoring for unauthorized access. It's not optional.
Most teams building agentic systems right now are focused on capability. Security should be the first layer, not the last. Excited to see this getting stage time at RSAC.
0
0
0
8