- While investors are spooked by talk of an AI-driven “SaaS‑Pocalypse,” analysts say telcos face no imminent software collapse
- Agentic AI will speed up modernization and automation
- Workforce impacts will hit repetitive, screen‑based roles first
While AI apocalyptic thinking is in the air, telcos shouldn't be preparing for the end of the world. At least not yet.
Wall Street panicked early this month, fearing that recent advances in AI would trigger the imminent obsolescence of enterprise Software-as-a-Service offerings from the likes of Salesforce, Workday and Oracle. Software, financial services and financial stocks lost $285 billion on Feb. 3 in an event nicknamed the "SaaS-Pocalypse."
The fear from investors is that enterprises — and telcos — will be able to vibe-code their own software. And SaaS companies aren't the only ones threatened: Microsoft AI CEO Mustafa Suleyman predicted that most or all white-collar tasks will be automated by AI with "human-level performance," within the next year to year-and-a-half. He cited lawyers, accountants, project management and marketing as vulnerable fields. That trend is already well under way in software engineering.
Like just before Covid hit
The current moment in AI is comparable to the weeks before the Covid pandemic, wrote Matt Shumer, CEO of OthersideAI, an applied AI company, in an essay titled "Something Big Is Happening" that went viral this month. The present moment in AI is "much, much bigger than Covid," he said. AI can already code better than humans, and now that it can do that, AI is making itself smarter.
Within one to five years — or less — AI will be better than humans at every job done by computers, Shumer said.
In other words: Everybody panic! Or, maybe not.
The sky is not falling
Shumer's essay was met with skepticism, some abrupt and pungent, others scoffing. "It's AI-generated word salad posted by someone with a vested interest in spreading AI hype," said one critic on X.
Meanwhile, industry analysts are skeptical of an imminent AI apocalypse within the telco industry. While AI is undoubtedly transformative, telcos and their software estates are large and complex, and transformation will take years to play out.
The SaaS-Pocalypse is overblown in the short term and oversimplifies what agents can accomplish, John Watson, Bell Canada group president, told Fierce. "That being said, we see big opportunity for agents to help accelerate app modernization and further cloud adoption," he said.
A gradual shift
AI agents can reduce switching costs for OSS/BSS, particularly legacy systems, but that replacement will be "layered, not wholesale," Roy Chua, AvidThink principal, said. AI will be able to automate tasks like ticket triage, order fallout resolution, swivel-chair operations across five different systems, and multi-step provisioning with known exception patterns, he said.
However, core OSS/BSS modules will be more resistant to replacement. "Full replacement of the deep, deterministic, compliance-heavy cores by generic agents? That's somewhat further down the line," he said.
And agentic AI can work to incumbent systems' advantage. "The same agents that make migration easier can also extend the life of installed monoliths by wrapping them effectively," he said.
Chua added, "So what I'd expect is gradual decomposition: agents and lightweight microservices growing around the core, individual modules getting swapped as contracts roll and risk tolerance increases, but the transactional core persisting for a long time."
Vendors make the transition
OSS/BSS vendors can see what's coming, and are wrapping OSS/BSS cores with agentic subsystems and AI orchestration to protect their business, Chua said.
Just this month, Amdocs launched a new agentic AI operating system (aOS), an open framework to let operators build agents in partnership with any vendors. Calix said 500 of its 1,500 ISP customers are now using the Calix One agentic AI platform. And Blue Planet launched its Cloud Native Platform for OSS in 2024, allowing telcos to plug in third-party AIs or roll their own, for BYO OSS AI.
Core systems like provisioning, routing, switching and billing will be augmented by AI, said Tekonyx President and Chief Research Officer Sid Nag. "Instead of rigid, workflow-driven processes, agentic systems will interpret intent," he said. "Systems will be able to call on APIs and optimize outcomes that drive more efficient operations and revenue management."
He added, "Over time, OSS/BSS systems become a sort of executive back end while the real control plane moves to an AI agentic-driven orchestration engine."
How to protect jobs
Telco workers can protect their jobs by retraining, Nag said. Workers most requiring retraining will be those that work in customer-facing operations, like customer care, taking calls on complaints, outages and billing errors.
Field operations will also be impacted. "You still need somebody to turn a screwdriver and put in physical fiber, but the diagnosis part will get automated," he said.
Ticket-driven and repetitive tasks mediated through screens, rather than requiring deep domain judgment or physical presence, are most susceptible to AI automation, Chua said. "Tier-1 contact center agents, especially chat, email and scripted voice, are at the top of the list; this is the telco equivalent of the L1 support roles being automated across every industry right now," he said.
Read next: What Fierce Network is watching for at MWC 2026: AI, 5G and network transformation