A clear majority of telecom operators believe that the development of large language models (LLMs) tailored to their specific generative AI needs would deliver significant benefits to the industry, argues a new study. According to a new report from TelecomTV, titled ‘AI-Native Telco Strategies,’ 72% of respondents were in favour of telecom-specific LLMs, highlighting a growing focus on sector-specific AI solutions.
Conducted in November 2024, the survey gathered responses from 64 participants across more than 45 telecom operators worldwide, with the exception of Latin America. Respondents’ roles ranged from CEOs and CTOs to AI-focused professionals, reflecting diverse perspectives across leadership and technical teams. While 72% expressed the need for tailored LLMs, 17% believed general-purpose LLMs developed by AI companies would meet telecom requirements, and 11% were uncertain.
Telecom sector shows strong commitment to AI adoption and strategy
The findings indicate an increasing emphasis on AI adoption within the telecom sector. According to the report, 70% of respondents said their organisations have established well-defined AI strategies, while 17% reported a lack of such plans, and 13% were unsure. On the leadership front, 76% confirmed their companies have appointed an executive specifically responsible for AI strategy, while 72% noted the existence of dedicated AI teams within their organisations.
AI’s influence on telecom operations also appears to be to growing. The survey revealed that 11% of respondents see AI impacting “everything” in their day-to-day operations, while 26% described its impact as “quite significant.” Another 39% indicated that AI is gradually changing an increasing number of processes, and 22% said AI affects only a few processes. Only 2% reported no impact on operations at all.
Revenue generation and cost reduction remain central to telco AI strategies. The report found that 86% of respondents believe AI can drive the development of new revenue-generating services. Similarly, 82% said AI has the potential to significantly reduce operational costs, reinforcing its role as a critical tool for business optimisation.
In terms of market opportunities, 63% of respondents identified the enterprise sector as the primary area for AI-driven growth, compared to 14% who pointed to consumer markets. This reflects ongoing efforts by telecom operators to expand their AI-enabled solutions in the enterprise segment.
Challenges related to AI adoption were also evident. 70% of respondents indicated that their organisations need to hire additional staff with proven AI development skills. However, the potential impact of AI on staffing levels remains uncertain. While 44% believe that AI-driven automation will lead to job reductions, 31% disagreed, and 25% were unsure of the outcome.
The debate surrounding telecom-specific LLMs has intensified as operators consider the value of customised AI solutions. Some companies, such as those within the Global Telco AI Alliance, are actively pursuing the development of LLMs for the telecom sector. However, during a recent briefing, Orange’s chief technology and information officer, Bruno Zerbib, and group CTO Laurent Leboucher said that the company remains unconvinced about the need for sector-specific models, though it continues to monitor developments.
The report also explored whether telecom operators are positioned to influence the broader AI ecosystem. 56% of respondents said telcos could lead innovation in AI development, while 38% believed operators would primarily serve as users of tools created by external AI providers.
When considering AI’s potential over the next three years, 92% of respondents expect significant improvements in network operations and management. Other key areas identified include network planning (70%), customer engagement (75%), energy efficiency (63%), and internal efficiencies (73%).
Meanwhile, in November, Nvidia announced that SoftBank’s Japanese telecom division would be the first to deploy its advanced Blackwell architecture chips, marking a significant step in the expansion of their partnership to enhance AI capabilities within Japan’s telecom industry. Looking ahead, the companies plan to integrate Nvidia’s Grace Blackwell platform into future systems. In addition, SoftBank has piloted the world’s first AI-powered 5G telecom network, leveraging Nvidia’s AI Aerial platform. This move is expected to open up major revenue opportunities for telecom operators, strengthening the role of AI in reshaping the telecom landscape.
Read more: Nvidia and SoftBank to drive Japan’s telecom transformation with Blackwell AI chips
