Telcos risk higher costs as AI adoption scales

AI agents and tokens could make up 30% of telcos' total costs

Telecom operators risk inflating operating costs rather than cutting them as artificial intelligence (AI) adoption scales, according to Bain & Company.

The consultancy said an "agentic operating model" is emerging in which traditional operating expenses make up approximately 70% to 80% of the total, whilst the remaining 20% to 30% consists of AI agent and token costs.

Bain warned that falling AI model prices do not automatically translate to lower costs.

“Although model prices have declined roughly tenfold each year, the effective cost per task often stays flat and the total token bill balloons unpredictably as usage rises,” it said.

It cited AT&T's experience redesigning its AI orchestration so large "super agents" delegated work to smaller, specialised models, which by the company's account reduced costs by up to 90% whilst tripling throughput.

Bain also flagged the risk of telcos deploying AI without redesigning workflows.

“Human-led operations don’t evolve, [Software as a Service] licenses continue growing, and AI simply becomes another line item on the income statement. Instead of transforming operations, organisations automate fragments of legacy complexity,” it said.

The consultancy also noted that telcos often focus on low-risk AI demonstrations that only marginally improve efficiency, which sets the ambition too low and misses the P&L upside.

“Instead of asking how AI can reduce the cost of existing work, leaders should consider which customer problems become economically viable only because AI exists. That shift changes the conversation completely,” it added.

The firm recommended five immediate actions, including measuring cost per resolved task, creating a dedicated AI compute budget, and eliminating wasteful AI usage.