Artificial Intelligence in Telecommunications 2025

3.2 Workforce, talent and culture As technical architecture and capabilities advance, required skill sets are shifting, with a decline in configuration tasks and an increase in software development. This has led to 64% of CSPs reporting a “high demand” for AI/ML specialists, with only 3% finding it “not difficult” to meet these demands.39 While it’s possible to reskill/upskill existing workers, one survey shows that only 21% of CSPs have a mature strategy for AI training or improving senior leadership’s fluency in AI concepts, technologies and applications.40 Cultural barriers compound talent gaps: 46% of CSP executives noted that talent is often aligned along business functions rather than broader capabilities, which stifles cross-functional collaboration, innovation and organizational agility.41 The scale of change across CSPs is significant. A recent analysis indicates that up to 65% of current working hours across functions will be transformed by LLMs (36% automated and 30% augmented).42 These statistics underscore a significant shift in the skills required and operational hours needed, challenging CSPs to manage this transition responsibly. Levels of automation and augmentation of worker activities within CSPs FIGURE 3 Higher potential for automation Higher potential for augmentation Lower potential for augmentation or automation Non-language tasksIT/technology Human resources Finance Customer sales and services Marketing Supply chain LegalOperations36% 39% 25% 1% 40% 29% 24% 7% 39% 31% 23% 7% 45% 24% 17% 15% 36% 30% 17% 17% 18% 43% 38% 1% 33% 19% 19% 29% 13% 30% 51% 6%4 million¹ On average, 65% of working hours can be transformed2 by LLMs. Of which, 36% of the time is susceptible to automation and 30% is susceptible to augmentation. It means that workers on average spend this amount of time on tasks that could be significantly impacted by genAl.Number of employees Note: Values may not add up to 100% due to rounding. 1. Total workforce included in the analysis. 2. Transformation is defined by the high potential for automation and augmentation. Source: Accenture research analysis based on O*NET and national statistical databases from 22 countries and 19,000 work tasks. CSP-specific data. 20 Artificial Intelligence in Telecommunications