Enterprises are confronting a "massive change" in how applications are used and data moves, according to Sullivan, who described the shift as a "revolution" driven by AI‑induced traffic spikes and new latency sensitivities.
He explained that early cloud adoption already altered traffic patterns, but the addition of scores of AI agents each week is further increasing volume and reshaping flows, creating a need for next‑generation networking that can operate at "machine speed".
Sullivan urged companies to assess internal skill gaps and consider specialist provider support, noting that a software‑defined architecture is essential for agility, cost reduction and rapid policy enforcement.
Research by Coevolve shows AI experimentation among organisations jumped from just over a third in 2025 to more than 70% today, yet the proportion that reports having appropriate AI governance rose only from 36% to 37%, indicating a readiness gap.
The CEO warned that without end‑to‑end visibility, organisations risk blind spots in user behaviour, potential data leakage and exposure to malicious use, especially as "bad actors" also gain access to AI tools.
To mitigate these risks, Sullivan advocated for zero‑trust models, continuous observability across distributed environments, and the ability to reroute traffic in real time when packet loss or "brownouts" occur.
He also highlighted organisational challenges, saying traditional silos between network and security teams must be broken down and staff re‑aligned to manage converging technologies effectively.
Finally, Sullivan noted a trend among large telcos to focus on specific markets rather than global connectivity, making interoperable, software‑defined infrastructure increasingly critical for stitching together services from multiple providers.