Ericsson adds ‘agentic AI’ to NetCloud enterprise assistant

Ericsson adds ‘agentic AI’ to NetCloud enterprise assistant is profiled by BTW Media because published evidence links it to internet infrastructure, governance, operational dependencies, or market visibility.

• ANA evolves from a prompt-driven GenAI chatbot to an “agentic” virtual expert with staged agents for troubleshooting, then configuration, deployment and policy.
• NetCloud will also support Ericsson Private 5G from Q4 2025, targeting adoption barriers with lifecycle management and AIOps insights.

What happened: Ericsson brings “agentic” workflow to network ops

Telecoms.com reports Ericsson has equipped NetCloud with what it calls the industry’s first enterprise 5G agentic AI virtual expert, enhancing the troubleshooting experience first, with further agents for configuration, deployment and policy slated through 2026. The upgrade moves ANA beyond simple prompts to intent-driven workflows and dynamic graphing of multi-point trends.

For context, Ericsson introduced ANA as a generative AI virtual expert in January 2025, designed to summarise documentation, assist configuration and speed diagnostics—now evolving toward a multi-agent model. The company has separately outlined how.

Also Read: NTT DATA unveils comprehensive agentic AI services
Also Read: Microsoft makes passkeys the default for new accounts

Why it’s important

If executed, agentic assistants could lower mean-time-to-repair and shrink the skills gap for lean IT/OT teams as private 5G expands. They also dovetail with Ericsson’s “talk to your network” direction and industry narratives about moving from copilots to more autonomous agents.

But open questions remain. How will enterprises govern agent actions, explain recommendations and manage model drift? What safeguards exist around data access and privacy when agents orchestrate tasks across WWAN and private 5G domains? Some experts warn that agentic systems raise fresh security and accountability concerns that must be addressed before wide adoption. A cautious, measured rollout—clear guardrails, audit trails and opt-in automation—will likely decide whether these pilots translate into mainstream operations.