Why Agentic AI Is Gaining Traction Among Global Telecoms Looking to Reduce Churn and Reignite Growth

The global telecommunications industry is at an inflection point. 

Over the last five years, the world’s largest communications companies have poured hundreds of billions of dollars into building out 5G networks. Most, however, have yet to fully realize the return on those investments. At the same time, telecoms have had to compete against a raft of new technologies and business models, from VoIP to fixed wireless to satellite services. 

This increasingly complex environment offers customers more choices than ever, leading to higher churn and lower revenues. Global telecommunications revenues are projected to grow a mere 3% this year — and just 1% for American carriers. Nearly a third of CEOs say boosting profitability is their number one challenge. 

“Today’s telecom companies are struggling,” said David Fan, Senior Vice President and General Manager for Communications, Media and Energy at Salesforce. “Hyper competition, market saturation, and disruptive business models are putting pressure on margins. They need new ways to reduce costs, drive revenues, and increase customer loyalty.”

The good news? Global communications companies are beginning to employ agentic AI solutions like Agentforce — ‌Salesforce’s digital labor platform — to streamline operations and deliver more personalized customer offerings. AI agents can perform tasks with minimal human intervention, streamlining processes and boosting productivity, while also increasing customer satisfaction. And while agentic AI won’t solve all of the industry’s challenges, it can uncover hidden efficiencies and unlock new revenue streams in ways no other technology can.

“Communications leaders are very excited about agentic AI,” Fan said. “They believe deploying digital labor to augment employees and automate tasks 24×7 can help improve both their top and bottom lines.”

Reducing costs by automating services

Agentic AI can serve telcos in many ways, but Fan is particularly excited about its potential to reduce the cost of interacting with customers while also helping internal teams ‌sell more services more efficiently.

For example, one of the most common reasons customers call into a telco’s contact center is when they have questions about their bill. Instead of making people wait for a support representative to come on the line and receiving answers from a static FAQ or, an AI agent can intelligently explain why their charges may be higher this month, help customers initiate a dispute, and offer ways to upgrade their service plans to avoid future issues. 

Agentic AI capabilities go beyond basic chatbots, which sometimes frustrate customers more than help them. AI agents can analyze all of a customer’s interactions with the telecom, offer personalized recommendations based on that data, and converse with them using human-like intelligence.

“By engaging customers in a conversational and empathetic way, autonomous agents can provide concierge-level service and make every person feel special,” Fan said.

Gartner predicts that autonomous agents will handle 80% of customer service interactions by 2029, reducing operational costs by 30%. Multiple studies have shown that using AI agents to resolve issues quickly increases satisfaction and reduces churn. At the same time, offloading most tasks to digital labor allows support teams to offer more personalized attention to the remaining 20%, enhancing customer loyalty. Researchers at McKinsey found that AI agents helped a large European telecom improve conversion rates for marketing campaigns by 40%, while also reducing its costs. 

Though we’re still in the early days of agentic AI, it’s already paying dividends for companies like One NZ. New Zealand’s leading telecom is using Agentforce to quickly deploy trusted AI agents in any workflow, answer customer questions, schedule appointments, help with plan upgrades, and use data to drive more personalized interactions. 

Customers like engaging with AI agents, too. According to Salesforce’s 7th annual State of the AI Connected Customer, more than a third of consumers and nearly half of business users say they’d prefer to work with an AI support agent if it meant getting faster service and avoiding unnecessary repetition. 

Boosting revenue by augmenting sales

In a similar vein, internally-facing AI agents can help drive revenue by acting like superintelligent sales assistants, providing deeper knowledge about product offerings, and making customized suggestions in real time. 

This is especially crucial in complex interactions such as B2B sales, said Fan. Instead of combing through product catalogs to see what offerings are available, a sales rep can have a conversation with an AI agent that understands each customer’s purchase history, knows what similar customers have ordered, and quickly suggests potential solutions and generates a quote. 

“With B2B, the services you sell are more complex and require a person to put together the right configurations,” said Fan. “Agents can help sellers assemble those configurations much more quickly. It’s a big game changer that will help them be much more productive.”

With B2B, the services you sell are more complex and require a person to put together the right configurations. Agents can help sellers assemble those configurations much more quickly. It’s a big game changer that will help them be much more productive.

David Fan, Senior Vice President and General Manager for Communications, Media and Energy at Salesforce

A major North American carrier, for example, is using Agentforce and Salesforce Data Cloud, Salesforce’s hyperscale data engine, to automate B2B sales quotes. 

“Without an effective data strategy, it’s going to be difficult for AI to do something magical,” said Fan. “You need to bring together all of your structured and unstructured data from any source and harmonize it to create unified customer profiles. Data Cloud is the trusted data source that powers every agent action, automation, and insight. And because Agentforce is built on Salesforce’s deeply unified platform, agents can access harmonized data from Data Cloud, plus industry-specific data leveraging the communications industry data model in Salesforce Communications Cloud.”  

By grounding Agentforce in industry-specific and personalized customer data, it can generate a quick and accurate quote based on product fit, serviceability, and pricing and discounting business rules set for the sales rep’s profile inside of Salesforce. This automation eliminates clicks, effort, and time, allowing reps to focus on deeper customer engagement and relationship building. This not only helps sellers increase revenue through more efficient selling, but also builds trust and loyalty with more personalized, useful quotes, ultimately reducing churn. 

“It’s important to help your B2B sellers understand both what they’re selling and what their customers need, so they can make the right offer to the right customer at the right time,” Fan adds. “That’s a great use case for agentic AI.”

Overcoming common misconceptions about AI agents

Still, a lot of companies have been hesitant to deploy AI agents because they view the technology as costly, complex, or hard to control. But the truth is that you don’t have to build your own large language model (LLM) to take advantage of agentic AI, said Fan. 

Agents can be deployed in a matter of days or weeks, thanks in part to Agentforce’s pre-built skills. Recently, at Mobile World Congress 2025, visitors to the Salesforce pavilion were able to build simple customer service agents in as little as 15 minutes that could answer questions based on a company’s website. For example, participants built agents for handling specific customer service use cases like billing inquiries, technical trouble shooting, product comparisons to quickly provide customers with information about available offers, pricing. 

The core services telecoms deliver don’t vary much from one company to another, Fan said. Every carrier needs to provide price quotes, resolve billing issues, offer plan upgrades, and schedule installation and repair services. These common telco agents can be easily customized for each carrier’s needs. 

But the biggest barriers to AI adoption remain trust and transparency, he added. Without robust governance systems in place, AI can put an organization at risk. 

“Many people fear AI is a black box, where you’re unable to understand how it works or predict what it’s going to do,” Fan said. “That is where Agentforce makes such a huge difference. Trust and transparency are the foundation for everything we do. Agentforce allows our customers to see everything their agents are doing and apply custom guardrails to limit the actions an agent may take.”

AI agents can achieve things that might have seemed like magic just a few short years ago. “Agentic AI is really about extending the workforce by having people and agents working together,” Fan said. “It’s about combining human creativity and AI-driven efficiency to unlock unprecedented levels of productivity and innovation.”

Go deeper: 

• Read the Telecommunications Guide to AI agents
• Learn how One NZ is building AI-enabled telecommunications
• Read how Bouygues Telecom is placing AI at the heart of customer relations