By: Tony Martino

Telecoms fraud protection and prevention is a strategic imperative for Communication Service Providers (CSPs). It safeguards revenue, trust, and customer loyalty whilst providing opportunities to differentiate in an increasingly competitive market. 

Fraud in telecoms is not a new problem, but the pace and complexity of threats has intensified. Fraudsters have adapted to the realities of hybrid working, cloud-based communications, and the age of AI. What has not changed is the scale of the challenge: the Communications Fraud Control Association (CFCA) estimates that telecoms fraud costs operators more than $38 billion USD annually in the US alone.

Fraud today takes two principal forms: toll fraud and scam calls. While both are distinct in nature, they are increasingly linked by shared tactics such as spoofing and automation. Together, they represent a profound challenge for the industry.

Fraudsters are becoming increasingly astute

Telephony networks, relied upon by billions, remain lucrative targets. As technology evolves, so too do the techniques used by criminals. Where in the past fraud was often opportunistic, today it is systematic and highly organised. 

Hybrid working practices and the migration to cloud platforms have widened the attack surface. Criminals exploit vulnerabilities in enterprise systems, SIP trunks, or poorly secured VoIP platforms to generate fraudulent traffic. At the same time, consumer-facing scams have been industrialised through robocalling platforms and AI-driven bots, allowing millions of calls to be placed in minutes. 

Toll fraud: a financial drain on operators and enterprises

Toll fraud is one of the most financially damaging categories of telecoms fraud. It typically exploits weaknesses in enterprise or carrier systems to generate revenue at the expense of legitimate users. A common method is Wangiri fraud, where victims receive a missed call from an international or premium-rate number and return the call, incurring high charges. 

International Revenue Share Fraud (IRSF) is another common tactic. Criminals compromise a phone system and generate thousands of calls to high-cost destinations in collusion with premium-rate number operators. Interconnect bypass fraud is when fraudsters reroute legitimate calls to avoid interconnect charges, and artificial traffic generation is when they inflate call volumes to create revenue share payments. 

The financial consequences of these activities are significant. Businesses can face bills running into tens of thousands of dollars within hours. Carriers, and any resellers or distributors in the supply chain, must then decide whether to absorb the loss or pass it on, often leading to disputes and reputational damage. 

Beyond direct cost, toll fraud undermines trust between providers, resellers, and customers. Enterprises expect protection as part of their service. When fraud occurs, blame is rarely directed at the criminals but rather at the provider who “allowed it to happen”.

Scam calls: a crisis of trust for consumers

Scam calls represent the other side of the equation. Unlike toll fraud, which exploits provider infrastructure for profit, scam calls directly target consumers. They are designed to manipulate, mislead, and defraud individuals. 

Examples include impersonation scams when fraudsters claim to be from a bank, government agency, or telecoms provider. Criminals also use investment scams to offer unrealistic returns to extract money from victims, or romance scams to exploit trust built in personal interactions. The scale of losses is striking. In the US, the Federal Trade Commission reported 5.7 billion lost to investment scams in 2023. In Australia, Scamwatch recorded nearly 20,000 scams in a single month in 2025, with financial losses exceeding $36 million AUD. 

The challenges with static rule-based systems

Fraud and scam tactics evolve rapidly, rendering static rule-based systems inadequate. Basic spend caps and call thresholds often detect issues too late, by which time the criminal is likely to have racked up many thousands of dollars of income and moved on to the next victim. 

These legacy systems also monitor call patterns to high-value destinations and use simple algorithms to trigger alerts when too many calls are made within a set timeframe. They often demand regular manual intervention, day and night. For