Innovate Americas: AT&T and Verizon spell out AI progress

TM Forum's Innovate Americas 2024 kicked off with TM Forum CTO, George Glass, sending the clear message that “AI is now the reality”. It was a message confirmed during the opening keynotes by AT&T’s SVP and Chief Data Officer, Andy Markus, and Verizon’s SVP Technology Strategy, Kamran Ziaee, who spelled out how AI usage is developing across their organizations.

TM Forum boosts AI collaboration

Glass noted that while many CSPs began by using AI components to improve or augment existing operations, in many cases AI systems were bolted onto the existing framework. The key step now, Glass said, is for CSPs is to “rethink our operating models for running AI end-to-end.”

Achieving network autonomy and its efficiencies is a key reason CSPs want to embed AI in operations, according to Glass: “You can’t reach autonomous network level 4” – the level where networks can self-configure and self-heal – “if you haven’t embedded AI into your design and operations processes.”

Doing so, Glass added, will require CSPs to “collaborate to imagine how a telco will work in the AI era.” TM Forum, he believes, has become the neutral ground where that collaboration will take place. To that end, Glass encourages CSPs to consider three core elements to drive new revenue growth with AI.

• Connect. While CSPs have always been in the business of connecting people, the existing operating models can also limit new revenue generation opportunities especially as CSPs adopt or shift to a connectivity-as-a-service model.
• Protect. Trust and security are crucial at every level with AI, from who has access to do what and whether the data and outputs themselves can be trusted.
• Create. Glass notes that the digital revolution has shown CSPs that “future growth will not come from our ideas entirely” but through collaboration across communities and organizations.

Glass noted that the TM Forum Open Digital Architecture (ODA) Canvas is designed to enable industry collaboration around AI and is now supported by all three major hyperscalers.

TM Forum, with AWS, has also launched GAMIT, a generative AI GenAI) maturity tool designed to help telcos benchmark their GenAI maturity against both regional peers and global leaders.

AT&T has integrated AI into its business

AT&T SVP and Chief Data Officer Andy Markus, who noted that AI is not new within AT&T because the company was “in the room when it happened” – referencing AT&T’s historic involvement in the development of traditional AI and machine learning (ML) systems.

The difference today, Markus explained, is that AI is now “integrated into how we run our business” with generative AI substantially expanding where the company can apply AI. . The objective is to “create value using these tools,” Markus said, which encourages a focus on use cases that will impact the top or bottom lines.

While traditional AI focused on “core blocking and tackling”, Markus explained, GenAI opened up AI access to a much larger group of employees who have been “waiting for this moment to capitalize on AI technology.”

As a result, said Markus, “we want to bring this to all parts of the company and help them do their jobs with reusable technology,” like Ask AT&T, a GenAI-based system that answers a range of operations and support questions for employees. .

Markus adds that there are four key aspects of Gen AI which AT&T focuses on: accuracy; cost; latency; and interoperability. Risk is also a factor, especially as AT&T invests in different technologies, because “tech is changing so fast there’s no guarantee the leaders today will be the leaders in three months,” Markus said.

Verizon focuses on what’s possible and practical with AI

We should all be tired of hearing about AI after the past two-year frenzy, quipped Verizon has also been working with AI technologies for a decade or more, but mainly to create “bespoke solutions to solve specific problems”, according to Verizon SVP Technology Strategy Kamran Ziaee. Ultimately,, however, this approach is not scalable. The goal instead has become “to industrialize AI across Verizon and have it be part of our DNA,” Ziaee said.

Ziaee explained that to make AI work as part of its business, Verizon focuses on two key elements:

• Having a very educated workforce that understands what’s possible with AI
• Empowering employees, such as with online training resources anyone can use to learn how to use AI in their jobs and make it part of Verizon’s fabric

From there, he says, Verizon can use AI to overcome fundamental challenges, like shifting from a household-based, customer model to one focused on individual customers; to support customers better with natural language support bots; and even to summarize emails and notes to help employees catch up on meetings or information they may have missed during the day an remain better prepared.

Ultimately, Ziaee believes AI will play a greater role in solving problems before they happen as well as helping Verizon to optimize its capital spend in the network.