HomeNews1Canada’s telecommunication sector utilizing AI to boost efficiency, but concerns about trust remains

Canada’s telecommunication sector utilizing AI to boost efficiency, but concerns about trust remains

Canada’s telecommunication sector utilizing AI to boost efficiency, but concerns about trust remains

Canada’s telecommunications sector is poised to ramp up its use of artificial intelligence to find ways to save costs, simplify customer experiences and increase revenues, industry watchers say.

 

The role of AI was a major theme at last month’s Canadian Telecom Summit in Toronto, where representatives from Canada’s major carriers, manufacturers and researchers converged for the annual three-day conference.

 

While responsible use of the technology remains an area of concern, attendees heard that telecom consumers as well as those who work in the sector are likely to increasingly encounter AI in the years to come.

 

Generative AI can help “transform” the sector at a time when companies are under immense financial pressure, said Hadi Skalli, managing director of data and AI for professional services firm Accenture. He said companies are facing challenges ranging from unfavourable regulatory decisions to increased customer expectations.

 

But he said Accenture’s research in global markets shows 46 per cent of all working hours in the sector can be either augmented or automated by generative AI. He said the company identified more than 90 use cases for telecom companies.

 

“AI is not new in the telco space, whether it’s in Canada or globally,” Skalli said.

“Telcos have been (using) AI for a long time … This is really about supercharging existing AI and adding gen AI on top of it.”

 

He said cost savings can be found by using AI to make customer service more efficient. For instance, a call centre backstopped by generative AI would be able to analyze phone calls in real time and coach agents on how to respond to customer questions or complaints much faster.

 

Skalli said that could also lead to opportunities to upsell customers thanks to a deeper understanding of their wants and needs.

 

“We’re starting to see various different telcos publish public facing chatbots to really help with customer experience in a way that traditional chatbots could never do,” he said.

 

“You felt like you were talking to a robot and I think none of us really want to do that.”

 

Companies such as Telus Corp. say they have already started integrating AI into their customer service operations.

 

Nazim Benhadid, chief technology officer for Telus, said a troubleshooting tool available for customers on the company’s website is now “completely gen AI.” It can read a problem outlined by the customer to assess their needs and determine how to resolve it.

 

Internally, he said Telus’ generative AI-powered IT service desk can perform similar tasks.

 

“We’ve seen through this that 50 per cent of the tickets are now closed automatically,” he said.

 

Benhadid said his company views the role of AI as being about “assistance,” noting it hopes to build on other developments in the telecom AI arena. That includes an announcement by Apple last month that it is launching Apple Intelligence, which is set to add generative AI features to its iPhone, iPad and Mac products.

 

“I think that’s going to kind of tell us a little bit about what people are ready for and where the value is,” he said.

 

Other uses for Telus include increasing sustainability — Benhadid said Telus has developed an AI framework that shuts down networks’ radio frequencies when their utilization is low to save energy consumption — and streamlining technicians’ workflow.

 

He said an assistant built for field workers that is integrated into Google Chat keeps them informed of their daily jobs, including driving directions, specific tasks they need to perform and instructions for how to do so.

 

But increased applications of AI have also raised concerns surrounding trust and privacy.

 

“There has to be a path that looks not only to leverage these technologies, but also to do so in the most responsible way possible,” said Sean Kennedy, who leads Nokia Bell Labs’ Artificial Intelligence Research Lab, at last month’s conference in Toronto.

 

Last September, the federal government launched a voluntary code of conduct for generative AI meant to quell anxiety over its proliferation and pace of development.

 

The code asks companies to agree to undertake several measures aimed at reducing the risks of AI, including screening data sets for potential biases and monitoring systems for potential harms. They also align with six key principles that include equity, transparency and human oversight.

 

“If we want to move from fear to opportunity, we need to build trust,” Industry Minister François-Philippe Champagne said in an interview during the summit.

 

Building that trust can take time, especially when AI disrupts the way tasks have long been conducted within the industry, said Piotr Wierzcholski, head of the telco OSS inventory subdivision at Comarch, a global IT company based in Poland.

 

Wierzcholski said at the conference that for engineers who have been building networks for decades, the introduction of new technology can be jarring.

 

“They’re building the network from scratch. They were here 20 years ago and when that level of automation is entering, they are feeling insecure,” he said.

“It’s part of the mindset change we need to all go through and that will not happen suddenly. That’s a process that needs to go step by step by going together with AI and learning how does it work.”

 

But Kennedy said there’s significant upside: generative AI has the potential to solve problems in the telecom sector better than humans can alone.

 

He said one of most practical uses is the development of industry-specific large language models. While many have grown familiar with LLM technology thanks to ChatGPT, he said that program falls short when it comes to understanding “the jargon that we use every day” in the telecom world.

 

Telecom-specific LLMs, on the other hand, are able to “take in specialized, focused knowledge and really synthesize these into answers that we think are meaningful,” according to Kennedy.

 

“Not only do they run more efficiently than humans and solve problems, but on top of that, they actually increase our ability to scale and build solutions more efficiently,” he said.

 

“They’re starting to actually have insights that we would say are human.”

 

 

 

 

This article was first reported by The Canadian Press