Cisco frames AI as a security and resilience priority, focusing on defending critical infrastructure with real-time, agentive controls and guardrails to distinguish benign from malicious AI behavior.

Patel notes frontier model training costs in the low billions and acknowledges a bubble-versus-secular-shift debate amid frothy valuations during rapid experimentation in AI markets.

The company emphasizes earning customer admiration daily through partnerships and competition, while pointing to partner-driven hardware advances, such as Nvidia’s optical switches.

AI is treated as a secular megatrend fueling demand, with aggressive plans for AI-enabled products, including a goal for 70% of software to be AI-written next year and a rapid transition away from legacy code within two years.

Cisco plans a substantial AI-focused investment, expanding R&D to about $8 billion and reallocating talent to high-demand areas like silicon and photonics, addressing workforce transitions with upfront reallocation rather than lay-offs tied to AI disruption.

Patel argues against halting AI development during the competitive race, advocating defender advantages and maintaining a deliberate gap between frontier and attacker capabilities to curb misuse while continuing innovation.

Cisco’s control over silicon, photonics, optics, and networking software gives it a vertically integrated edge and strengthens supply-chain resilience.

The company positions itself as a cornerstone of AI infrastructure, thanks to its integrated hardware, software, and silicon stack that enable AI workloads and robust data-center connectivity.