Cisco bets on networking-security combo to capture $28B AI infrastructure opportunity

Cisco unveiled a comprehensive AI strategy at Cisco Live 2025, positioning the company to leverage its unique combination of networking, security, and silicon capabilities for enterprise AI deployments. CEO Chuck Robbins emphasized that Cisco’s dual expertise in networking and security creates a competitive advantage that neither pure networking nor security companies can match, particularly crucial as agentic AI requires constant high-level activity unlike traditional generative AI chatbots.

The big picture: Cisco is betting that the infrastructure transformation needed for AI will mirror the networking boom of the 1990s that originally made the company a juggernaut.

• Robbins noted that 85% of enterprises believe they must “do AI” in the next 18 months, creating massive demand for AI-ready infrastructure.
• The company announced 19 major datacenter innovations in just the past six months, demonstrating accelerated product development.
• Chief Product Officer Jeetu Patel predicted AI agents will make “8 billion [people] feel like 80 billion” through productivity gains, but warned of constraints in power, networking, and compute.

Key infrastructure announcements: Cisco launched several platforms designed to simplify AI operations while maintaining security.

• The unified Nexus Dashboard creates “one brain for all of our data center fabrics,” according to Patel.
• AgenticOps platform includes AI Canvas, a multiplayer management console that provides real-time infrastructure visibility and automated remediation.
• Hybrid Mesh Firewall enables distributed policy enforcement and works with existing third-party firewalls.
• Cisco Live Protect can shield networks from exploits “within minutes” compared to the industry average of 45 days to patch vulnerabilities.

Strategic partnerships and integrations: Cisco deepened its collaboration with key AI players to strengthen its platform approach.

• Cisco switches are completely integrated into Nvidia architecture, with Nvidia NeMo models secured by Cisco AI Defense.
• The $28 billion Splunk acquisition enables data correlation across multiple domains to improve digital resilience.
• Catalyst switches are now unified with the Meraki network platform through a single management dashboard.

What they’re saying: Industry leaders highlighted the fundamental role of networking in AI success.

• “None of our network friends have security and none of our security friends have networking,” Robbins said, emphasizing Cisco’s unique position.
• “If people don’t trust the system, they’re not going to use it,” Patel explained regarding security as a prerequisite for enterprise AI adoption.
• “The way in which you run your network will never be the same again,” Patel said about the AgenticOps platform’s impact.

Product development transformation: Cisco has restructured its development process to accelerate innovation.

• Most new AI software was developed by small teams of six to eight members, representing a cultural shift from the company’s traditional approach.
• The company launched “one of the largest refreshes of networking devices in Cisco history,” including smart switches with isolated compute capabilities.
• New devices act as sensors providing environmental data back to the system for enhanced observability.

Market positioning challenges: Despite impressive technical capabilities, Cisco faces messaging and differentiation hurdles.

• The company needs to better communicate its Silicon One chip initiative, launched in 2019, given current AI chip company valuations.
• While Cisco claims AI is “foundational” in its products, this messaging has become common among enterprise vendors.
• The sheer volume of announcements creates a challenge in helping customers understand specific payoffs across different product areas.