- Telcos are newly excited about AI, especially for speeding up modernization
- AI code generation and early chatbot capabilities were oversold, leading to skepticism
- New telco leadership is pushing cultural change to break down silos and enable meaningful AI implementation
It’s been three and a half years since ChatGPT burst onto the scene, kicking off the AI era and sending a ripple of excitement across the globe. That enthusiasm waned during a recent bout of disillusion, but it seems telecom operators are once again excited about what AI can do for them, AWS telco chief Jan Hofmeyr told Fierce.
According to Hofmeyr, the use case generating the most buzz these days is AI-enabled modernization – think mainframe and VMware migrations – and for good reason.
“The challenge with mainframe was it was such a difficult modernization effort that takes many years, most probably more than one CIO lifetime, to do these migrations and the business case just never made sense,” Hofmeyr said. So, he added, operators just extended the life of these systems for as long as possible. The difference now is AI can help.
In February, AT&T inked a multi-pronged deal with AWS to, among other things, migrate critical IT and operational business support applications to on-premises AWS Outposts environments using AWS Transform.
Updated with an agentic AI toolset in December, AWS Transform is specifically designed to speed modernization of old enterprise code and applications. Hofmeyr said Transform takes care of everything from creating documentation for existing systems to making a migration plan based on dependencies and conducting testing and validation. What used to take years can now be completed in a much shorter timeframe, he said.
“Every CEO, CTO meeting I had [at Mobile World Congress] wanted to go see that,” Hofmeyr said. Though the telecom industry is still heavily based on-premises, Hofmeyr said he believes Transform will “unlock that movement” to cloud environments. AT&T, for instance, will be migrating to AWS Outposts.
“I think they are stuck on premise because of the complexity – the perceived complexity – of their applications and systems,” he said. “I really think that this is most probably the pivotal moment where they now see a path to migrate these applications and modernize them.”
Trough of disillusionment
But Hofmeyr said he understands why operators and enterprises at large may now be skeptical of what AI promises to do for them.
For instance, Hofmeyr said he believes the transformative power of AI code generation was oversold. The problem, he explained, is not that AI does a bad job at code generation. It’s that there are many more steps involved in getting code into production than just writing it. So, using AI code generation without updating any of the other processes involved just ends up “filling up the funnel,” he said.
Hofmeyr added a lot of enterprises – especially telcos – are still clinging to the chat agents they deployed early in their AI journeys. These, however, don’t offer the kind of transformation companies are looking for.
“It streamlines, but real transformation happens when systems become AI-enabled, when you start actually using agents within your organization that really do meaningful things, real critical work for you,” he said. Hofmeyr argued recent updates to AWS AgentCore have given companies the governance and security tools needed to do this without giving their CTO and CISOs a heart attack.
Forging ahead with fresh blood
One of the biggest AI-related hurdles telcos face is cultural resistance. A recent Fierce Network Research survey report found nearly a quarter of respondents said cultural barriers were their biggest obstacle in turning AI into a growth engine. Lumen Technologies and MetTel, which were both interviewed for the report, detailed some of the challenges they faced in changing internal mindsets.
Hofmeyr noted a number of telcos now have new leadership that has brought a “very different mindset that is breaking down these barriers” and eliminating siloed operations. He pointed to BT, AT&T, T-Mobile and Verizon as examples.
“Every attempt to break down those siloes has pretty much failed and I think it’s deeply rooted in the culture,” he said. “I’m starting to see leadership coming into telco that have absolutely zero tolerance for that and are driving the right outcomes…I really feel there’s a shift.”