- Telecom operators are struggling to move beyond isolated AI deployments and scale efforts organization-wide
- Entities like TM Forum, GSMA and Altman Solon are offering governance frameworks to help telcos navigate AI adoption
- Clean data access and transparency in AI usage are just some elements they say are critical for AI success
Roadmaps, pillars, frameworks—call them what you want, telcos need some hand-holding when it comes to their artificial intelligence (AI) strategies. Fortunately, a growing list of organizations have been happy to oblige with their own guides on how to implement AI responsibly and effectively.
“There is nearly universal recognition that AI will lead to a paradigm shift in telco’s operating models. However, very few operators have a clear picture of what that operating model will or should be,” George Glass, CTO of TM Forum told Fierce Network.
According to Glass, while AI promises to revolutionize telcos’ operating models, many operators have only implemented isolated use cases. "Most AI deployments today are ‘bolt-on’ use cases – specific, isolated deployments - rather than ones that represent part of an overall architecture and operating model," he said. Areas like customer experience, network planning and energy savings have already seen some success, but the industry is still struggling to scale AI efforts across their organizations.
While it’s clear that AI presents an opportunity to change the trajectory of the global industry and “build a path back to growth,” he said this can only be realized by greater cooperation and collaboration across the tech and telco ecosystem.
For its part, TM Forum released a report, “Building An AI Strategy,” in which leading operators shared their insights and challenges in implementing AI.
Alix Jagueneau, head of external affairs at industry group GSMA, which represents the interests of mobile network operators, noted that clear policies and consistent regulations are critical for the sector to fully realize AI’s potential. Specifically, she emphasized the need for a harmonized policy environment.
“Policymakers and regulators can facilitate and stimulate further research and development investment in AI in both the public and private sectors,” Jagueneau told Fierce. “Making potential use cases economically viable is another important consideration, as return on investment is crucial in securing further funding.”
Data governance is critical
One of the tallest hurdles telcos face is access to good data, a prerequisite for AI success. “Access to clean, quality, usable data is the single biggest challenge we are hearing from our members,” Glass said.
Both TM Forum and GSMA’s frameworks are addressing this through AI-ready data architectures and guidelines for responsible AI use. The GSMA’s Responsible AI Maturity Roadmap gives telcos a framework to roll out AI across their services in “evolving, ethical, consistent and sustainable ways.”
Transparency remains a critical factor in ensuring AI’s responsible use, Jagueneau said. It is essential to be transparent about when an AI system is being used, what kind of data it uses and its purpose. This comes from designing, developing and deploying AI in a responsible and ethical way that is “human-centric and rights-oriented.”
For example, she noted Telefónica is telling its customers what type of data is used for AI systems, how it is used and when they are interacting with an AI system.
Priya Mehra, director at consulting firm Altman Solon, noted even telcos and CSPs that have been using traditional AI for years are “realizing that the traditional AI and data governance is not fully adequate to manage generative AI.”
And of course, companies with less mature AI “have an even bigger pill to swallow,” she told Fierce.
Beyond data
Altman Solon’s “Generative AI Governance Frameworks for Telecommunications Leaders” covers AI data strategy, but Mehra noted that AI governance needs to be thought of “as a much broader area than just data governance.”
As such, Altman Solon outlined a four-part governance framework for telcos:
- Strategy and operating guidelines: Defining the vision and establishing guardrails.
- Organizational structure: Setting up governance and ethics committees to ensure oversight and accountability.
- Processes: Implementing traceability, access controls, risk checklists and continuous model monitoring.
- Technology: Using platforms to monitor biases, accuracy and performance of AI models.
Mehra said in addition to a governance framework, telcos must also invest in internal AI expertise to fully leverage these tools. Employees need to know how to use them efficiently and responsibly.
She concluded, “Telcos are figuring out ‘how do I protect myself from liability? How do I protect my brand reputation and image?’” Because at the end of the day, there’s a lot at stake.