Australia’s telecommunications industry is set to follow the global telecom market in adopting generative AI use cases, as local providers face the prospect of losing ground to competitors in areas such as productivity and customer service if they don’t invest in the new technology.

A survey from Amazon Web Services found global telcos are embracing generative AI for customer service chatbots and employee assistance tools. A number of local telcos are already using AI tools, and generative AI could enhance competition with over-the-top providers.

Jump to:

How are global telcos approaching generative AI technology?

The AWS survey, conducted by Altman Solon, polled 100 senior telco leaders from the U.S., Western Europe and the Asia-Pacific, including Australasia. Testing 17 use cases across marketing and product, customer service, network and IT, it found telcos are already investing in generative AI, particularly to improve productivity and customer service.

Generative AI is already being implemented globally

The AWS survey found global telcos have started implementing generative AI. On average, each use case tested had a 19% adoption rate, indicating implementation had started or was being planned. This adoption rate was expected to grow to 48% within two years.

In addition, telcos expected spending on generative AI to increase up to six times in two years, with 45% of respondents saying spending will rise to between 2%–6% of total tech spend, up from 1% today.

Generative AI applications are distinct from traditional AI

70% of telcos surveyed see the incremental value served by generative AI as distinct and significant from existing AI and machine learning. Sixty-four percent agreed most use cases are new applications, not served by existing nongenerative AI applications and processes.

SEE: Read more here about how generative AI really works.

Chatbots are the most popular generative AI use case

Customer chatbots are the most widely adopted use case, with 63% already in production. Productivity is another focus, with use cases including employee assistance with contact center documentation or network operations knowledge management.

North America leads generative AI adoption

The North American market is leading adoption with an average use case adoption of 21%, while APAC as a whole was lagging slightly at 16%. AWS put this down to the more limited capabilities of existing generative AI models in non-English languages.

Australian telcos are also joining the generative AI race

Australian telecommunications companies will be compelled to follow the global market.

“There will be no choice but to adopt generative AI to remain competitive,” said Anton Gain, managing director at Gain IT & T Consulting. “Everyone is talking about generative AI, and how it can help with enhanced customer service, network optimization, security and fraud detection and predictive network maintenance.”

Gain, who assists Australian business and government clients with their telecommunications infrastructure, told TechRepublic that there is no doubt that AI will form a ” … greater and greater role in running a telco in the future.”

However, he said that, at present, much of the conversation is conceptual in nature; though, there is one use case of generative AI that is expected to surface first.

“My experience is that customer service productivity through chatbots and voice AI is the first user case being tackled,” Gain said.

Australian telecommunications players already exploiting AI tools

Louise Hyland, CEO of the Australian Mobile Telecommunications Association, told TechRepublic that Australian telcos have a track record of adopting AI technologies and tools.

For example, Optus has partnered with mobile voice recording and AI platform Dubber, by adding their meeting and call transcription service, automating vital call monitoring and recordkeeping.

SEE: Explore our comprehensive artificial intelligence cheat sheet.

“That helps with adherence to banking and finance regulations,” Hyland said. “Generative AI then provides the opportunity to use the data gathered to train a personalized large language model, which can then effortlessly produce communications tailored to the organization.”

Other telcos are making similar moves. Last year, Telstra hired Orla Glynn as the organization’s executive responsible for AI and automation. The telco has already been using AI to assess whether a text message is a scam and stop it from reaching end customers.

“Ericsson and TPG have been trialling a cloud-native, AI-powered analytics tool that provides insights about the operator’s 4G and 5G subscriber base,” Hyland said. “It uses ‘smart data collection with embedded intelligence’ to predict and resolve performance issues in real time.”

“Australian telcos have embraced the use of AI technology,” Hyland said. “Over this decade, national adoption of 5G will be a key enabler for the success of artificial intelligence and realization of its economic, social and environmental benefits for Australian communities.”

Local telcos will face barriers in generative AI adoption

Gain IT & T Consulting’s Gain said he expects local telcos to take a cautious approach to adopting the technology.

“There are stringent regulatory, privacy and data security barriers in Australia that need to be considered and overcome before implementing generative AI solutions,” Gain said.

This reflects the view of many telcos around the world. The AWS survey found that 61% of surveyed telcos indicated they had concerns around data security, privacy and governance.

“For telcos to leverage generative AI for company purposes, it requires a large set of proprietary data,” said Ishwar Parulkar, global chief technologist for telecommunications at AWS. “While there are many public LLMs, there is concern that proprietary company data could be embedded into the public model itself, creating intellectual property risk.”

Industry skills gap to drive off-the-shelf generative AI model uptake

Telcos also face in-house and broader industry generative AI technical skills gaps. Some telcos cited their lack of technical resources as a barrier to generative AI adoption, with only 15% of surveyed telcos indicating a desire to build foundation models in-house.

The rest expected to use off-the-shelf models. However about two-thirds (65%) of respondents anticipate that they will train those same off-the-shelf models with proprietary internal data to tailor them to their specific needs.

Future market leaders could be made with generative AI

How successfully telcos adapt to generative AI could impact the future market. Parulkar said those telcos who move to embrace generative AI would be able to compete more effectively with other organizations like over-the-top providers who have “taken over the value chain.”

“The industry is looking at generative AI because it is really open season for anybody who wants to get in and learn about it now — we are just starting to scratch the surface,” said Parulkar. “Anyone who gets into it and explores how it can help the top-line could emerge as a winner.”