AI calling services took a significant step forward on March 13, 2026, as GSMA published a white paper at the 5G Future Networks Summit during the Mobile World Congress (MWC) 2026 in Barcelona, Spain. The paper, titled “Gigabit Uplink Networks, Deterministic Latency, and Network Evolution for the Mobile AI Era,” outlines development trends, use cases, and business models for native voice communication services in the mobile AI era.
White Paper Defines AI Calling Services Framework
The white paper states that the integration of 5G-A networks and artificial intelligence has moved mobile communications into the mobile AI era. Telecom operators are transitioning native voice services from traditional calls to AI-powered voice calls. By embedding AI algorithms and computing capabilities into the native IMS-based voice network, traditional voice calls are evolving into enhanced services and new applications.
Key AI Calling Applications Identified
The paper identifies AI noise reduction as a representative application of immersive AI calling. AI algorithms remove ambient noise across multiple environments, including offices with noise levels above 40 dB, streets above 60 dB, and construction sites above 80 dB. This allows users to access high-quality voice services without relying on device hardware capabilities.
AI-powered real-time translation represents another key application under interactive AI calling. The paper states that language barriers can be progressively eliminated through enhanced voice network capabilities. The service provides real-time transcription and accurate translation during video calls, supporting business professionals in international online conferences, travelers abroad, and individuals with hearing impairments.
New Business Model for Telecom Operators
The white paper highlights that telecom operators can integrate AI capabilities into native voice services to develop new business models. Users can access AI-enhanced functions during traditional calls after subscribing to the service. Consequently, operators can shift from generating revenue through single-traffic connectivity to generating revenue through a multi-dimensional experience model.
Six-Indicator User Experience Evaluation Model
The paper systematically defines user experience evaluation model specifications for AI calling. In addition to the three existing indicators — Quality of Experience (QoE), Quality of Service (QoS), and coverage — three new indicators are added: immersive AI calling experience, interactive AI calling experience, and Quality of Intelligence (QoI). Moreover, immersive calling can significantly improve Mean Opinion Score (MOS) and Signal-to-Noise Ratio (SNR) in core voice calls.
Interactive calling requires the network to support new interactive channels and capabilities, including a Data Channel (DC) and a Video Channel (VC). These channels enable advanced experiences such as screen sharing, real-time translation, and interaction with AI agents. QoI serves as a key indicator for measuring the intelligence level of the voice communication network, covering high-quality AI model usage, flexible AI management, and network and user state awareness.
Industry Collaboration and Standardization Outlook
The International Telecommunication Union (ITU) has launched a work item called P.AI-MOS to evaluate user experiences in multimodal AI applications. However, proposals for AI calling experience standards remain under study. GSMA and its industry partners are calling for joint efforts to establish rules linking Key Quality Indicators (KQIs) for AI applications with network Key Performance Indicators (KPIs).
These efforts aim to accelerate the standardization of AI service experience for mobile networks. The initiative is intended to provide stronger support for the broader development of the mobile AI industry, as telecommunications networks continue to evolve toward intelligence-first architectures.