Telcos evaluate choices on the path to becoming AI-native

Earlier this month the CEOs of BT and Vodafone each gave different reasons for making significant cuts to their workforces. But taken together the separate announcements highlight some of the questions facing the telecoms industry as it weighs up how to benefit from AI and automation. Namely, will cash-strapped telco businesses that have been struggling to grow revenues for the last decade be able to invest in AI to innovate in services that deliver a competitive advantage? Or will they have to wield AI primarily as a cost reduction tool?

Vodafone’s new CEO, Margherita Della Valle, placed the decision to cut a total of 11,000 posts from an employee base of 90,000 within the context of the company’s poor return on capital investment in Europe.

BT’s CEO Philip Jansen, meanwhile, who expects the company to shed a total of 55,000 jobs by 2030 from today’s workforce of 130,000, positioned his decision in the light of investment in innovation: Much of the reduction in BT’s employee numbers, he said, will arise from it completing the roll out of more efficient fiber and 5G access networks and the decommissioning of its copper network.

What grabbed the headlines, however, was what Jansen had to say about the company’s use of AI: “For a company like BT there is a huge opportunity to use AI to be more efficient … There is a sort of 10,000 reduction from that sort of automated digitization. We will be a huge beneficiary of AI.”

AI productivity gains

BT is not alone in expecting a lot from AI. Ahmed Hafez, VP Technology Strategy, Deutsche Telekom, speaking on a panel about AI at Telecom TV’s DSP Leaders World Forum, sees AI “as … the biggest transformation … we will ever encounter. And this is not only about the magnitude of what AI will do, but also … the pace.” Citing a study by Accenture, which found that AI would increase business productivity by at least 40%, Hafez said: “I would think this is very much an understatement. 40% is too low. I expect it to be much higher than that.”

At Vodafone, velocity is the biggest change that automation had brought to date, according to its chief technology officer, Scott Petty, also speaking at the Telecom TV event, adding that Vodafone’s largest adoption of automation currently is in operations.

Going forward, however, greater automation means telcos will have to scale up their use of AI beyond a few dozen separate use cases. And it promises to pose a major challenge for which many telcos are not ready. “We will not have tens of use cases we will have hundreds or thousands of use cases,” said Hafez. “Are we ready to embrace such scale? Are we building AI for scale? I don't think so.”

In the meantime, other tech companies have moved much earlier and much faster to deploy AI, noted Joe Butler, Chief Technology Officer, Digital Catapult during the Telecom TV panel discussion. “Tech companies [use of] AI … is moving at breakneck speed, and business opportunities and value creation … [are] going astonishingly quickly,” according to Butler. For example, “between 2013 and 2016, Google's implementation of AI throughout their company was completely exponential,” he said. “So, when we talk about missing the boat on AI in telco, we have …[done] already. We should be clear about that.”

Investing in growth

Meanwhile Neil McRae, Chief Network Strategist, Juniper Networks, and until recently, chief architect at BT, encouraged telcos to focus on using AI to deliver “double-digit growth”, rather than cutting costs. He warned that if the industry doesn’t then “assume you won't be able to afford AI. We won't be generating enough money to actually build it properly.”

Telcos' ability to invest heavily in innovation, however, has been constrained by financial performance. Commenting on Vodafone’s announcement of job reductions during TowerXchange Meetup Europe 2023 conference, Brian Potterill, Director of Competition Policy at Ofcom pointed out that: “If the go-to strategy is cost-cutting it is not a sign of a healthy industry,” adding that ongoing investment by telcos depends on the industry being sufficiently profitable.

McRae would also like to see CSPs using AI to gain an edge over their competitors, without getting bogged down in defining frameworks or getting caught up in regulation.

But telcos that have to juggle the complexity of operating legacy systems and networks while introducing new technologies seem likely to continue working together.

Deutsche Telekom’s Ahmed believes that “we need some common basics …[and] principles that we align to at a high level, but which doesn’t slow us down. Some common understanding about how AI will evolve.” Deutsche Telekom, for example, would like to see collaboration on access to real time data. “We tend not to have data in the moment universally across our operation. We tend to have [it] 15 minutes later ... so, time and commonality are two very good fundamental building blocks [of] ... good collaboration for the industry.”

Telcos will also need to think about how to allocate their AI resources effectively.

“You're not solving one small problem,” pointed out Azfar Aslam, Chief Technology Officer Europe, Nokia, during the panel discussion on AI at DSP Leaders World Forum. “You're looking at a whole domain and bringing it all together," he explained. “In real businesses, you are competing for capital to solve a problem there are millions of problems … and that capital is always finite," explained Aslam. "As the cost of implementing AI starts to go up, you have to be really sharp about which problem you choose to solve first.”