From Risk to Resilience: a GMS Network Protection Webinar Overview

The recent GMS network protection webinar looked ahead to 2026, focusing on how operators can defend revenues and trust in a landscape shaped by OTT bypass, flash calls, AI-powered fraud and rapidly growing data traffic. The event was framed as a conversation among three perspectives — commercial, operational, and R&D — around a single question: how to keep A2P messaging and network protection relevant, effective, and profitable in the coming years.

The agenda guided participants from industry challenges and customer success stories to AI-powered protection and technical foundations, with live polls used to take the pulse of the audience.

Commercial Reality: A2P Revenues Under Pressure 

From a commercial perspective, the webinar described an A2P market that has been a strong revenue source for almost a decade but now faces serious headwinds. Forecasts presented during the session point to a significant dent of around 16 billion USD in A2P SMS revenues by 2025, even as other parts of the messaging ecosystem grow towards 19 billion USD.

The pressure is not just on volume but on integrity. SMS fraud continues to rise, with fraud patterns becoming more sophisticated. Voice fraud, especially voice bypass, has been growing since the COVID period. Flash calls and robocalls were described as part of a “new reality” the industry must address, with flash call authentications projected to reach around 23 billion.

Over the next five years, operators are expected to lose around 1.3 billion USD in A2P revenues if nothing changes. Enterprises are drawn to alternatives such as flash calls because they are far cheaper than SMS at a time when SMS prices are described as having gone “crazy high”. The webinar framed the challenge as a need to manage this situation in totality rather than fight it channel by channel.

Poll Insights: What Keeps A2P Relevant? 

A live poll asked participants which strategy they considered most effective for keeping international A2P SMS relevant for authentication and revenue over the long term. All options were presented as valid, but the results revealed a clear preference pattern.

There was a tie between securing long-term deals with enterprises while maintaining current rates and strengthening network protection against OTT bypass, including the introduction of data firewalls. The least favoured option was a drop in international rates, reflecting an apparent reluctance to resolve the problem through price cuts.

The outcome underscored that operators are looking for a combination of commercial strategy and stronger technical protection, rather than discounting their way out of structural changes in the ecosystem.

Operational Insights: AI-Powered Grey Traffic 

From an operational standpoint, the webinar described a “silent surge” of AI-powered grey traffic: spam SMS, spam calls, robocalls and various forms of unauthorised traffic. What makes this wave particularly challenging is the use of AI to mimic legitimate traffic, blurring the line between real and fraudulent events and allowing attackers to bypass traditional, rule-based tools whenever loopholes exist.

The impact is threefold: revenue loss for operators, erosion of user trust and increased risk of regulatory penalties. The session framed real-time AI and machine-learning-based network protection as essential rather than optional.

AI was shown to reshape traffic filtering by automatically grouping similar messages, upgrading analytics by templating fraud patterns, accelerating signature analysis, improving prediction through consistent trend evaluation, and transforming traffic correlation from reactive recovery to real-time distinction between legitimate and bypassed flows.

Four Pillars of Protection  

The GMS network protection suite was presented as spanning four domains: messaging, voice, signalling and data. Messaging protection centres on a tier-one SMS firewall with proactive spam detection, signature-based analysis, auto-blocking and 24/7 monitoring. Voice protection uses AI and ML to profile and validate calls, detect and block flash calls and robocalls, and forecast voice fraud.

On the signalling side, the solution is being extended to SS7, Diameter, and GTP, with GSMA-compliant rules designed to withstand cross-protocol attacks. For data, it can detect A2P, RCS and OTT traffic, disrupt alternative channels and push traffic back to SMS where needed, supported by signature fingerprinting.

A closer look at the SMS firewall showed four core blocks: monitoring with flexible metrics, KPIs and alerts; analytics that verify SMS signatures, group by senders and use AI/ML for pattern detection; blocking driven by ad-hoc policies and advanced robotic process automation for real-time response; and advanced features including external APIs, SMS honeypots and full backup and restore capabilities. The firewall not only detects and blocks threats but also continuously improves the messaging ecosystem's security posture. 

R&D insights: deep dive into SMS Firewall architecture  

TThe R&D segment focused on design principles and product evolution. Core qualities include low latency in high-throughput environments, such as 200,000 messages per second at peaks, where processing must stay in the low millisecond range; flexible fraud detection that can combine multiple criteria; a fraud base built from previous deployments; and constant improvement via feedback from live traffic into configuration, signatures and machine-learning models.

Recent evolution in 2025 includes a blocking API that gives end users and operations teams a straightforward way to block specific numbers or groups, and performance optimisations that moved one customer case from a 97% throughput rate (with 3% drops) to full throughput with an average processing time of about 20 milliseconds.

For 2026, the roadmap includes a unified API to simplify integrations, centralised management so multi-site deployments can be configured from a single graphical interface, and an AI-based anti-smishing tool capable of deep inspection of text and URLs to stop more malicious SMS content.

A high-level technical overview described how messages are decoded from protocols such as HTTP, SCTP, and TCP, passed through flexible rule sets and actions, then re-encoded into formats such as M3UA, MAP, and TCAP, so that the firewall remains transparent to the surrounding infrastructure. Kafka, an SMS analysis component and Elasticsearch form the machine-learning pipeline: traffic is sampled asynchronously, signatures are built and stored, and each new message is checked against the latest signatures before being allowed or blocked.

Throughout this part of the webinar, simplicity, reliability, and future-proofing were presented as guiding principles, with the solution positioned as a “future SMS firewall” built to address both current and emerging fraud patterns.

In closing, the webinar emphasised that this session is not the end of the conversation. We encouraged our participants to stay connected through the GMS website, where detailed information about products and solutions is available, and to subscribe to the newsletter or contact us directly for more insights, case studies, and updates.