Globe Telecom quantifies AI ROI through operational integration

Globe Telecom has realized P1.2 billion in total business benefits from its artificial intelligence initiatives as of August 2026, marking a transition toward rigorous financial accountability for machine learning deployments. The company reports a 2.8 times return on its cumulative AI investments, encompassing revenue generation, cost avoidance, and productivity improvements across its enterprise stack.

Incremental revenue accounted for the largest share of these gains, contributing over P825 million to the total. This capital inflow stems from AI-driven personalization engines integrated into Globe Rewards, roaming services, lending platforms, and postpaid subscription management, which allow for more precise customer targeting and service delivery.

Cost avoidance initiatives yielded an additional P368 million in benefits by optimizing workflows in finance, marketing technology, and order management. These applications demonstrate the utility of AI in addressing specific operational bottlenecks rather than serving as general-purpose automation tools, effectively reducing the overhead associated with manual data processing.

Productivity gains reached P28 million, derived from 52,000 hours of labor reclaimed through the automation of repetitive tasks. This shift allows personnel to reallocate time toward high-judgment activities, effectively increasing the output capacity of existing teams without expanding headcount or increasing infrastructure complexity.

The organization achieved a 90 percent adoption rate among its workforce by August, supported by a growing internal community of over 450 AI builders. This decentralized approach enables non-technical units to develop domain-specific solutions, reducing the reliance on centralized IT bottlenecks for minor process improvements and accelerating the deployment of localized automation scripts.

Globe has integrated these models directly into existing business processes to address specific operational challenges, ensuring that the technology serves as a functional layer within the current stack. By focusing on integration rather than standalone experimentation, the company has minimized the friction often associated with adopting new machine learning frameworks in large-scale telecommunications environments.

The company confirmed that the reported P1.2 billion figure represents a composite of revenue, savings, and productivity rather than net profit. This distinction is critical for enterprise architects who must distinguish between gross operational efficiency and the bottom-line impact of long-term AI maintenance and compute costs.

Carl Cruz, president and chief executive officer at Globe, emphasized that the firm is prioritizing the measurement of value creation over mere adoption metrics. This strategy reflects a broader enterprise shift toward validating AI efficacy through tangible financial indicators and operational performance benchmarks, moving away from the speculative hype that often characterizes early-stage technology adoption.

Technical leaders should observe how Globe balances this decentralized builder model with the necessity for centralized governance. As the organization moves beyond initial integration, the primary challenge remains maintaining the precision of these measurement frameworks as AI complexity scales within the infrastructure, ensuring that data integrity and security posture are not compromised by the rapid proliferation of internal tools.

The governance of these AI systems requires a consistent framework to ensure that the 450-plus internal builders adhere to established security and data privacy standards. Without such oversight, the risk of technical debt and shadow IT increases, potentially undermining the long-term architectural integrity of the organization’s data flow.

Future performance will depend on the sustainability of these gains as the company refines its tracking mechanisms to account for the total cost of ownership. Stakeholders should monitor whether these productivity improvements translate into sustained margin expansion or if the costs of maintaining these models, including data storage and model retraining, eventually offset the initial efficiency dividends observed in the current fiscal period.