Based on Telefónica’s Alex Harmand, agentic AI adjustments how community APIs generate income by shifting pricing away from particular person API calls
Telefónica and Nokia have just lately collaborated to check how AI software program brokers may assist using community APIs, as a part of the GSMA Open Gateway initiative to standardize entry to telecom community capabilities. Preliminary testing has targeted on a fraud prevention use case in a lab setting. Telefónica is utilizing Nokia’s Community Publicity Platform to reveal community APIs and associated features — together with SIM swap and system swap — whereas Nokia’s Community as Code platform aggregates these capabilities for utility builders.
The collaboration displays a broader trade shift towards agentic AI, the place autonomous software program can uncover, mix, and monetize community capabilities with out direct developer orchestration.
RCR Wi-fi Information caught up with Telefónica’s Core and Platforms Senior Supervisor Alex Harmand to debate the highest use instances — past fraud detection — that can drive probably the most significant income from agent-enabled community APIs.
Right here’s what he stated:
Identification and Belief Orchestration. An AI agent can dynamically combine quantity verification, SIM standing, system context, and placement information to generate a real-time contextual identification confidence rating. “This helps onboarding, account restoration, and step-up authentication with clear monetization by way of decreased fraud losses and decrease friction,” stated Harmand.
Context-aware threat scoring for funds and commerce. On this use case, the agent evaluates transaction context in actual time and selects probably the most related belief alerts dynamically, relatively than counting on static fraud guidelines, which Harmand stated permits “adaptive authentication flows and risk-based pricing fashions” that straight enhance approval charges and cut back chargebacks for fee suppliers and e-commerce platforms.
Predictive Intelligence-as-a-Service. Prediction-driven brokers transfer past easy validation towards proactive optimization. At MWC, Telefónica and Mavenir will showcase a No-Present Prediction agent that forecasts attendance chance and triggers focused interventions to scale back missed appointments. Based on Harmand, this use case has clear monetization potential throughout healthcare, retail, area providers, and logistics, the place no-shows straight impression income.
Community high quality and SLA optimization for enterprises. For B2B prospects, an AI agent may interpret intent — similar to making certain optimum connectivity for a particular utility — and dynamically consider components like latency, congestion, slice availability, and coverage constraints. Whereas nonetheless early-stage, it represents a promising future use case.
“That is significantly related for cloud gaming, media streaming, IoT fleets, and mission-critical enterprise providers,” stated Harmand, including that this income mannequin is tied to premium connectivity tiers, SLA-backed providers, and dynamic quality-on-demand choices.
From promoting technical entry to promoting outcomes
Huge image, defined Harmand, agentic AI adjustments how community APIs generate income by shifting pricing away from particular person API calls towards outcome-based consumption. As an alternative of billing per endpoint request, operators monetize every agent execution — which means prospects pay for a accomplished goal, similar to a fraud evaluation, identification verification, or no-show prediction.
Behind the scenes, an agent could orchestrate a number of APIs and community capabilities, however commercially it’s bought as a single, productized motion. This simplifies pricing whereas permitting operators to seize better worth from higher-level automation.
The mannequin additionally permits operators to bundle capabilities into bundled providers relatively than exposing standalone APIs. Presents similar to Identification & Belief or Commerce Threat Scoring mix a number of alerts and features into one answer, creating clearer industrial propositions for enterprises.
In apply, agentic APIs shift monetization from promoting technical entry to promoting outcomes — enhancing buyer predictability whereas rising common income per consumer by pricing the answer itself relatively than every underlying microservice.
“The shopper nonetheless will get a predictable industrial mannequin, and we seize extra worth as a result of one agent run could orchestrate a number of underlying capabilities,” stated Harmand.
