HPE's networking boss says AI will handle all trouble tickets without humans in two years

What happened

A senior HPE networking executive has predicted that artificial intelligence will be capable of resolving all network trouble tickets without human intervention within two years, according to a report by The Register. The executive argued that IT teams will come to trust autonomous AI agents to diagnose and fix network issues in much the same way drivers stopped worrying about how automatic transmissions work under the bonnet.

The comments position HPE as making a confident, time-bound bet on agentic AI replacing manual triage in network operations centres, rather than simply assisting human engineers. The claim was made in the context of HPE's networking business and its broader push to embed AI-driven automation into enterprise IT infrastructure management.

Why it matters

This is a statement about where enterprise IT operations are heading, not where they are today. If AI agents can genuinely triage, diagnose and close trouble tickets end-to-end, it reshapes the operating model for network and IT service management: fewer tier-1 and tier-2 support roles, faster resolution times, and a shift in human focus toward exception-handling and oversight rather than routine fixes.

The "automatic transmission" analogy is doing a lot of work here. It frames trust in autonomous systems as something that builds gradually through reliability, not through understanding the mechanics. For technology leaders, the real question isn't whether AI can technically resolve tickets — vendors have been promising this for years — but whether organisations are willing to remove humans from the loop before trust, auditability and failure-mode transparency catch up with the technology's capability.

The Renascence take

Predictions like this tend to be judged on the technology, but the harder problem is behavioural, not technical.

Nobody trusts an automatic gearbox because they understand the engineering — they trust it because decades of consistent, low-stakes repetition made the risk invisible. Network operations don't have that luxury: a handful of visible, high-cost failures early on will do more to shape adoption than any roadmap promise. Operators serious about this shift should spend less time benchmarking ticket-resolution accuracy and more time designing the escalation paths, audit trails and "it's okay to be wrong sometimes" governance that make staff and customers comfortable ceding control — because trust in autonomous IT, like trust in autonomous transmissions, will be earned in increments, not declared on a two-year timeline.