AI and modernization for telecom transformation: PwC

Telecom operators are under pressure from rising costs, customers who expect instant digital service, and competitors that can launch faster. Many operators assume they must finish large-scale modernization before AI can deliver value. In practice, that sequence is too slow. The same legacy complexity that blocks transformation is also where AI can help first.

A better approach runs AI and modernization in parallel. Well-scoped AI agents can take on targeted work—like confirming orders or triaging tickets—while highlighting the exact bottlenecks to fix next. When agents repeatedly stall or escalate, it often points to broken handoffs, conflicting rules, missing data, or fragile integrations.

At the same time, you need to see through the hype. AI shouldn’t be deployed everywhere by default. If a rule, script, or standard automation can solve the problem reliably, that’s often the right first step. The real challenge is explaining AI’s value in a way that business stakeholders can trust, a way that’s tied to measurable outcomes, not novelty.

The economics add urgency. Data traffic keeps growing, but revenue isn’t keeping pace, tightening margins and making efficiency gains a top priority. Digital-first businesses are also resetting expectations for speed, personalization, and transparency.

AI agents can absorb high-volume, repeatable work across operations—freeing your employees to focus on judgment and exceptions. The result is faster resolution and more consistent execution across domains that are often managed in silos.

AI’s value becomes clearer when operators separate three sources of impact: improving an individual’s productivity (for example, faster investigation and better decision support), enhancing an end-to-end workflow (fewer handoffs and less rework), and enabling deeper analysis and innovation (pattern discovery across logs, tickets, and usage). Agents can also accelerate workflow breakdown and monitoring, identify process failures in the real world, and surface near-term standard operating procedure changes. And they can do this well before a full transformation is complete.

A data-first approach can help your telecom avoid two common missteps: automating broken processes and waiting years for a “clean slate.” Instead of building more point solutions on top of siloed applications and fragmented databases, you can start shifting toward an architecture where data and decisioning sit at the center. Over time, some specialized tools—like niche billing add-ons or heavily customized order systems—can be simplified or retired as agents operate across domains with shared context, consistent policies, and standard interfaces. Your enterprise stack moves from merely recording what happened to detecting what’s happening and triggering the next right action in near real time.

Why telecom operators can’t afford to wait

Many operators still carry decades of technology and process debt; things like fragmented product catalogs, duplicate billing platforms, brittle order flows, and data that’s inconsistent across systems. This makes even small changes slow, expensive, and difficult to predict. Traditional transformation programs that insist on end-to-end modernization before meaningful AI adoption underestimate how quickly the market is shifting.

The reality is simple. Waiting for a perfect core delays impact and creates room for faster-moving competitors. Telecom needs a model that balances stability with speed to help deliver benefits now while steadily removing the constraints that keep AI and the business from scaling.

In 2026, PwC predicts a majority of businesses will move from experimenting with AI to scaling it as a core operational capability. For telecoms, the path forward is a dual-track model—deploy AI now and modernize the specific parts of the business and technology stack that agents consistently trip over.

What to build: A dual-track model for the AI-native telecom

Use agents as both delivery tools and discovery tools. As agents execute defined tasks, they reveal where work breaks down—through repeated escalations, frequent fallbacks, missing context, or stalled actions. Those patterns typically trace back to inconsistent data, weak integrations, unclear policies, or unnecessary steps. Capturing and analyzing those signals helps teams pinpoint the modernization moves that can unlock greater value.

AI agents and people work best as a coordinated system with clear role boundaries. People set intent, define guardrails, and handle edge cases where judgment matters. Agents operate within those limits, coordinating steps across domains and escalating when confidence is low. To scale, telecoms should have updated roles, workflows, and training so teams can supervise agent behavior and manage exceptions.

Track 1: Start now with AI agents that can create immediate impact

Start with workflows where agents can clearly reduce cost or improve the experience, and you can measure success. High-impact areas include software engineering productivity, network and IT assurance, order capture and orchestration across complex services, and high-volume customer contact drivers such as bundling and pricing. These use cases can deliver measurable operating improvements when they’re designed with clear outcomes, reliable data access, and a controlled path to production.

Examples by domain illustrate the range of near-term value. In consumer operations, agents can identify churn risk and trigger tailored outreach. In fixed broadband, they can analyze serviceability and installation patterns to reduce missed appointments. In mobility, they can detect potential fraud in SIM-swap or porting flows and prompt verification. And in enterprise services, they can help build and validate complex orders. Importantly, agents can coordinate actions across domains that siloed teams often struggle to manage consistently.

Track 2: Use agent signals to guide focused modernization

Use what agents reveal to modernize business support systems (BSS), operations support systems (OSS), and enterprise support layers that slow down delivery and decision-making. PwC’s Agent Powered Performance research shows that AI agents can deliver significantly more impact, and cost less to run when they sit on top of a simplified, modern digital core of standard processes, a strong data foundation, and fewer legacy constraints.

A modern, ERP-backed core can give agents the added stability, traceability, and control points they need to act safely across the enterprise. As you build confidence with early agent deployments and can point to real results, that’s when you should accelerate deeper modernization—guided by evidence, not ideology.

The modern core, defined for executives

A modern telecom core helps give both people and agents the confidence to act safely, consistently, and at speed. It is not a single platform. It’s a set of capabilities that can reduce fragmentation and make decisions repeatable.

Four capabilities that matter:

Simplified commercial and charging backbone
One trusted product catalog with unified pricing and fewer billing variants, so offers are consistent; changes are faster, and cost to serve drops.
Composable BSS and OSS
Modular business and operations systems with well-defined APIs and event-driven integration, so upgrades are safer, and integrations are easier to change.
Governed data domains
Clearly owned customer, product, order, usage, and financial data—all treated as assets with defined quality, lineage, and access controls.
Closed-loop control points
Standard decision and action patterns for fallout, credit, collections, and assurance that give agents trusted “actuators” to execute within policy at scale.

Success story: Cutting costs and retaining customers

One operator used AI-supported taxonomy, automated data classification, and end-to-end lineage (tracked through an observability layer) to reduce data-operations costs by about 50%. Teams spent less time searching for root causes and more time assessing impact and fixing issues that mattered.

In practice, that meant fewer break-fix cycles, cleaner inputs to financial and customer reporting, and faster, better-governed launches of converged bundles—helping retain subscribers and increase share of wallet.

Modernization isn’t going away, but it is becoming more focused. The goal is to expand what AI agents can handle safely and reliably, so they can unlock real efficiency and better outcomes across the business. In other words, modernize selectively to remove friction and then let agents take on more of the work.

Modernization scorecard:
Where progress should show up

Modernization matters, but how do you know if it works? This scorecard can help you assess progress against five priorities from PwC’s 2025 TMT Innovation and Investment Readiness Survey—showing where ambitions align, where constraints persist, and where impact should become visible in day-to-day operations.

Priority area	What leaders are saying	What to watch for
Enhancing customer experience and engagement	51% of telecom executives rank customer experience and engagement among their top-3 strategic priorities for the next 12-18 months	Higher satisfaction, fewer complaints, more proactive resolutions
Sustainable cost savings and margin improvement	35% rank sustainable cost savings and margin improvement among their organization’s top-3 strategic priorities for the next 12-18 months	Lower service costs, faster outage resolution, AI-led routine requests
Speed and agility	38% cite lack of speed and agility as an inhibitor to capturing value in a dynamic business environment	New products launched in weeks, faster customer updates
Data readiness	27% cite insufficient data or insights as an inhibitor to navigating challenges and capturing value	Decisions based on trusted, unified data—not siloed or fragmented sets
Simplification	49% cite siloed organizational structures (such as poor cross-functional collaboration or misaligned goals) as an inhibitor to navigating challenges and capturing value	Fewer handoffs, decommissioned systems, seamless cross-team workflows

_{Qs: What are your organization’s top 3 strategic priorities over the next 12-18 months? (Rank top 3.) What are the biggest inhibitors your organization faces when navigating challenges and capturing value in a dynamic business environment? (Select up to 3.)

Base: Telecommunication execs 37

Source: PwC’s 2025 TMT Innovation and Investment Readiness Survey, September 2025}

AI agents introduce a new layer of digital labor

AI agents go beyond traditional automation because they can interpret context from multiple sources, monitor live conditions, and take bounded actions like flagging billing anomalies, assessing serviceability, or updating orders when conditions change. Unlike rules-only automation, agents can handle more variability, while still operating inside defined limits. When designed well, this reduces cross-team dependencies, speeds up resolution, and lowers cost-to-serve—especially in high-volume areas like customer care and network operations.

Reinvention comes from redesigning value streams

Reinvention with AI comes from redesigning whole value streams—from onboarding to assurance to billing—so work can continue with fewer stops and manual checks. Early agent deployments help because they make breakdowns visible. If agents fail due to conflicts in addressing data and serviceability rules, that becomes a clear modernization target: unify the decision logic.

As foundations improve, agents can take on broader responsibilities, moving toward safe, policy-based “zero-touch” execution where appropriate, while people stay in control of exceptions and risk.

What it takes to operate as a human and agent enterprise

Agents can change how work gets done. Decisions happen faster, collaboration happens in real time, and improvement can be continuous rather than episodic. To capture that value, you should have a practical operating model that includes new roles, clear ownership, and governance that enables speed without creating avoidable risk.

Decide where agents lead and where humans stay central

A practical way to plan deployment is along the telecom value chain: plan, build, run, serve, and sell. This aligns with the TM Forum eTOM model and helps leaders decide where humans should remain central, where agents can lead, and where tight coordination is required. The aim is not maximum automation. It’s the right allocation of responsibility to improve resilience and outcomes.

Orchestrate to make agents act as one

Agents deliver more value when they operate as a connected system, sharing context, rules, and authority. A coordination layer like PwC’s agent OS can help agents across customer, billing, network, and service domains act consistently. With enterprise intent and policies built in, a single network event can trigger aligned actions, including customer messaging, billing adjustments, and assurance steps. The result is fewer handoffs, faster resolution, and more confidence in semi-autonomous execution.ustomer messaging, billing adjustments, and assurance steps. The result is fewer handoffs, faster resolution, and more confidence in semi-autonomous execution.

Establish clear ownership to move beyond pilots

Scaling requires enterprise accountability. Whether ownership sits with a chief AI officer or a broader transformation leader, the mandate should extend beyond pilots to include agent strategy, data and technology foundations, workforce transition, and Responsible AI controls. Leadership also needs to manage the trade-offs that appear at scale—speed versus governance, standardization versus differentiation, central control versus domain autonomy.

Put guardrails in place to scale safely

A central-federated governance model can help your telecom move quickly without unnecessary risk. Shared standards for agent authority, data access, safety checks, and lineage provide consistency while domain teams retain ownership to innovate locally. Without guardrails, agents may act on outdated data, duplicated logic can drive inconsistent decisions, and automation can outpace human oversight during outages. Predictable governance keeps behavior consistent—even under pressure.

Design for scale from day one

Many AI efforts stall because scale is treated as an afterthought. Each agent should be tied to a business domain with a named owner, clear outcomes, explicit guardrails, and a roadmap to production. Focus internal build on agent logic that truly differentiates the operator—like assurance, retention, and complex order handling—and rely on collaborators for more standard capabilities. The goal is to move fast while keeping strategic control. Build where it matters. Buy where it doesn’t.

Move faster with PwC’s reference models for AI-native telecoms

Telecom transformation often stalls in long pilot cycles. PwC’s telecom reference model helps operators move faster by starting from how telecom runs today—grounded in more than 650 Level 4 eTOM processes—and enriching them with automation and agent design patterns.

From there, we map future-state flows using customer journeys, solution playbooks, and proven patterns—connecting today’s operations to an AI-ready model without rebuilding everything from scratch.

Our telecom reference model also draws on a library of over 3,000 AI use cases and deployments to help teams scope and design workflows tied to measurable outcomes.

Turning AI execution into competitive edge

Operators won’t lead by announcing AI ambitions or launching scattered pilots. They can win by executing in the places that move the P&L and the experience.

Deploy agents in high-value workflows.
Use agent friction to prioritize targeted modernization.
Modernize selectively to expand safe agent autonomy.
Redesign roles and incentives for a human-plus-agent workforce.
Coordinate agents as an enterprise system—not disconnected tools.

Reinvention with AI is hard work. It demands a clear-eyed view of how the business runs—the workflows, data, and constraints that have accumulated over time. It also requires a cultural shift that rewards new ways of working, equips teams to collaborate with agents, and treats AI as an operational capability rather than a side project.

For leaders who treat agents as the engine of near-term performance and modernization as the flywheel that increases what agents can handle, the payoff is concrete—faster execution, lower cost to serve, and a stronger position in a market that won’t wait.

The dual-track telecom transformation

Why telecom operators can’t afford to wait

What to build: A dual-track model for the AI-native telecom

Track 1: Start now with AI agents that can create immediate impact