- 47 percent of telecom AI deployments focus on customer care, with three-quarters motivated by internal efficiency and cost reduction.
- Asia leads in deployment volume, while Europe and North America show higher maturity—more than half of AI projects live rather than just trialled.
What happened: Nearly half of telecoms’ AI deployments target customer care, driven by cost savings and efficiency.
A recent GSMA Intelligence report, Telco AI: State of the Market, Q2 2025, reveals that 47 percent of tracked telecom artificial-intelligence (AI) initiatives centre on customer-care contexts, notably call centres and sales touchpoints.
It further reveals that 60 percent of these AI deployments are live and integrated into daily operations, while the remaining 40 percent remain in trial or planning stages. The report observes that, unlike traditional telco roll-outs like 3G/4G/5G—which unfold over a linear decade—AI often progresses through “cycles within a cycle”, repeating trials and roll-outs in a more dynamic fashion.
Crucially, 75–80 percent of deployments are driven by internal efficiency and cost-saving goals, compared to only 10–20 percent aimed at generating new business. Geographically, while Asia Pacific leads in sheer deployment numbers (averaging 3.9 AI projects per operator), Europe and North America excel in maturity, with 60–75 percent of initiatives already live.
Also read: Optus partners with Nokia to deploy cloud-native 5G voice services
Also read: Nokia drives Medusa cable linking Europe and Africa
Why it’s important
The telecom sector is clearly prioritising AI where it can deliver tangible savings—automating customer-care workflows to reduce staffing costs and churn, rather than chasing new revenue streams. This pragmatic use of AI in customer careunderscores a broader shift towards streamlining operations rather than pursuing flashy innovation.
The contrast between Asia Pacific’s volume and Europe/North America’s deployment maturity suggests strategic differences. Asia may be experimenting aggressively to push scale quickly; Western telcos, perhaps under greater regulatory and reputational scrutiny, favour cautious but stable live roll-outs.
While AI powers cost-efficiency, it also compels deeper consideration of its impact on the customer-agent experience. Reports from vendors like Cisco forecast that by 2028, 68 percent of service interactions may be handled by agentic AI—but still emphasise that human connection remains vital. This trend points to a future where telcos must blend AI efficiency with human empathy, rather than replace it.
Operators must also confront infrastructure readiness. A survey by Ciena and Heavy Reading found just 16 percent of telcos believe their optical networks are “very ready” for AI workloads, with 39 percent “ready but with work to do”, and 40 percent only “somewhat ready”. Efficiency gains via customer-care AI are only sustainable if the underlying networks can support increased AI traffic and latency demands.
