Vox executives unveil how real-time, on-device AI, semantic traffic analysis, and flexible anti-fraud infrastructure are driving a global shift in telecom security — from monetizing A2P in high-risk markets to preparing for threats like voice cloning and quantum disruption.
“When you start stopping the fraudsters, they adapt — so you need to adapt faster,” says Vox Solutions VP of Engineering, Alan Murphy. Artificial intelligence can help a lot here: “AI is now smart enough to know not just what a message says — but why it’s being sent.” That’s crucial in unmasking criminals.
The battle against fraud in messaging and voice is entering a new phase — one where generative AI, real-time detection, and even on-device decision-making are rewriting the playbook. Fraudsters are getting faster, smarter, and more evasive. The future is one where your phone knows what matters to you — and blocks what doesn’t — before you even see it.
There are signals of a major shift in the global messaging ecosystem. Some of them are coming from the team at Vox Solutions, fresh off a MEFFYS win in the Antifraud category — the Oscars of the mobile world. In an exclusive interview with the MEF podcast Perspectives, Alan Murphy and the company’s Chief of Staff, Teodor Mugureanu, offer insights into how AI is moving from the data center to your device, why SMS isn’t going anywhere, and what needs to change. And yes… there’s a hint of big news still under wraps: “We have a lot of news… but nothing I can disclose today,” says Mugureanu.
This is the future of trust in telecom, as seen from the top. Let’s dive in…
Alan, how would you characterize the current state of the international voice and messaging market?
Addressing telecom fraud starts with visibility — having the infrastructure to monitor messages and enforce anti-fraud policies. Vox and others in the industry provide systems that analyze network traffic, detect fraudulent patterns, and block malicious messages.
Fraudsters adapt quickly, so it’s crucial that operators continuously monitor threats and update defenses. Historically, this required human intervention. Now, with generative AI, we can automate and improve these processes. Our Clara platform enables operators to run detection systems more autonomously, reducing reliance on manual labor and continuous retraining — resulting in a more secure and scalable environment.
In markets with rampant fraud but limited regulatory oversight, how do you apply your tools?
Messaging and voice firewalls are critical. While regulators can’t stop fraud directly, they can enforce compliance. In regions lacking oversight, Vox helps by offering flexible commercial models that address both capital expenditure and operational challenges. We enable monetization of legitimate traffic while blocking fraudulent messages and enforcing policy compliance.
Teodor, Vox recently closed a deal in Jamaica tied to localized AI. What’s the broader role of AI in shaping telecom in these markets?
Each country approaches AI differently. Some, like the UAE, have well-defined strategies. Others are still developing regulatory frameworks. We’re ready to support local telecoms with AI infrastructure, though full implementation depends on the maturity of local policies.
From a technical view, how do AI and edge computing enhance fraud detection, Alan?
Quantum computing poses future encryption threats, but AI and edge computing offer immediate benefits. Eventually, users may have localized AI on their phones that understands context — deciding, for instance, whether a message from “the gym” is spam.
Currently, our AI-enhanced firewalls provide semantic analysis of traffic in real time, enabling rapid identification and neutralization of new fraud patterns. These tools evolve faster than the attackers can adapt — although attackers also have access to powerful technologies like voice cloning, which we are preparing for.
Telecom fraud costs Africa $1.5 billion annually. Is real progress being made, Teodor?
Yes, though progress varies. Africa’s A2P SMS business is critical — some operators derive up to 20% of their profits from it, compared to 1–3% in Europe. Larger operators are making solid strides in fraud prevention; we expect smaller ones to follow as technology adoption increases.
With programmable infrastructure and the cloud, what shifts do you foresee in voice and messaging delivery?
Initiatives like Open Gateway and RCS are evolving, but adoption is slower than expected. In the next 3–5 years, we’ll see: continued growth of OTT (over-the-top) platforms; RCS establishing itself in mature markets; AI simplifying how users interact with brands — e.g., AI assistants handling bookings or complaints. However, API-driven innovation like network slicing for gaming will only succeed with broad operator coverage — ideally 80–90% of a given market.
What should operators focus on in the coming months?
In terms of risks: Operators must stay competitive. A2P SMS isn’t a monopoly — it competes with IP-based channels. Falling behind means losing value.
In terms of opportunities: A2P SMS has untapped potential. Instead of being a pure utility, it can evolve into a targeted advertising channel, powered by data and operator insights. This opens new monetization avenues, especially for local enterprises and national markets.
