How Mobile Phone Networks are Adopting AI
Our mobile phones are getting smarter, but what about the telecom networks they run on?
Last week, Apple became the latest mobile phone company to announce the integration of artificial intelligence (AI) into its handset operating system. Called Apple Intelligence, it aims to make phones easier and faster to use by enhancing Siri, transforming it into an even more personal assistant. This follows similar moves by Samsung’s Galaxy AI and Google’s Gemini AI for their Pixel handsets.
The increase in AI usage means that phones will perform more computing, generating and using significantly more data. This puts additional pressure on mobile phone networks like O2, EE, Vodafone, and Three in the UK.
Ian Fogg, Director of Network Innovation at research consultancy CCS Insight, explains that to cope with this, telecom companies are also rapidly adopting AI. “Network operators are using AI to dynamically manage radio frequencies to provide optimal service levels and manage cell towers to use less energy during low demand times,” he says.
The increased use of AI for mobile phone network maintenance is now a global phenomenon. Alex Sinclair, CTO of GSMA, which represents global mobile operators, says that in South Korea, Korea Telecom can now localize and fix faults within a minute thanks to AI-enabled network monitoring. Meanwhile, in the US, AT&T uses predictive AI algorithms on trillions of network alerts to warn of potential issues.
Ian Fogg notes that network providers are increasingly using AI to monitor their systems. Vodafone and other operators are using AI digital twins – virtual replicas of real-world equipment like masts and antennas – to continuously monitor network performance. AI is also being used to manage how large data centers use energy to cool their servers and optimize storage capacity.
The explosion of data generated by increased AI usage is another reason why telecom companies worldwide are investing in so-called 5G standalone mobile networks. Unlike older, less efficient 4G systems, these rely on new, dedicated 5G infrastructure, offering much greater speed and capacity. However, some experts believe even this high-tech solution may not be sufficient to meet the demands of the AI era. For instance, at this year’s Mobile World Congress in Barcelona, some experts argued that AI won’t reach its full potential until the launch of 6G in 2028.
While mobile customers usually only notice networks when things go wrong, they become more aware when customer service is poor, potentially damaging brand reputations. Therefore, the industry also hopes that AI can fundamentally improve how it interacts with and serves customers.
For example, the Global Telco AI Alliance – a joint venture of Deutsche Telekom, e&, Singtel, SoftBank, and SK Telecom, serving 1.3 billion customers across 50 countries – aims to develop AI chatbots tailored specifically to the telecom sector and the types of questions customers typically ask. The founders hope these specially trained chatbots will handle most basic customer inquiries, freeing call center staff to focus on more complex issues.
Vodafone, meanwhile, is working with Microsoft’s Azure OpenAI service to improve customer service through its digital assistant, Toby, which interacts with over 40 million customers per month in 13 countries and 15 languages. The more questions Toby can answer without human intervention, the fewer complaints there are likely to be, which is good for profits and corporate reputation.
Scott Petty, Vodafone’s CTO, says, “We see AI primarily as a ‘virtual assistant’ for humans.” He argues that AI frees customer agents to work on more complex issues and helps employees find what customers are looking for on their websites more easily. Early evidence suggests that the number of successful online customer journeys has doubled, and call volumes have decreased by 10%.
While many commentators worry that AI could lead to mass job losses in the telecom sector as routine tasks are increasingly automated, GSMA’s Alex Sinclair believes it could be empowering, especially for low-income countries. He argues, “AI will provide a specialist tool to help emerging markets catch up.” He supports democratizing AI usage so that it is not just the preserve of the wealthy.
Ian Fogg of CCS Insight shares this optimistic outlook: “AI has been around for some years, used for specific [telecom] cases. But now, it is being applied in many more areas – networks, devices, and software – enhancing the potential of our devices.