In 2025, the number of fraudulent messages continued to rise, with North America and Europe experiencing the most significant increases. Phishing remained the most prevalent fraud category, while gambling and regulatory content also posed significant threats. SMS continued to be the dominant channel for fraudulent activity.

These are among the key findings of a new report by Infobip, an IT and telecommunications company based in Croatia. The report analyzed billions of interactions globally, highlighting a year of record volumes of blocked fraudulent traffic, the rapid scaling of intelligent, AI-powered defenses, and the persistent role of mobile phones as a key channel for fraud.

Fraud detection grows 77%

Fraud detection saw a remarkable growth of 77% from 2024 to 2025, reflecting both an increase in threat activity and the expansion of detection capabilities.

All detection methods experienced an increase. Pattern-based detection, which identifies harmful messages through content signatures, keyword matching, and traffic pattern analysis, saw a staggering 105% year-over-year (YoY) increase in blocked messages. URL detection, which identifies messages containing malicious or fraudulent links by scanning embedded URLs in real time, increased by 31% YoY. Finally, AI-powered detection, which uses text and image detection to identify harmful content, rose 71%.

AI-based image analysis, in particular, has become a critical layer of fraud detection. In this method, instead of placing suspicious text directly in messages, fraudsters hide harmful content inside images to bypass text-based filters. In 2025, image-detected blocks by Infobip increased by a remarkable sevenfold, implying that this technique is now used at high volumes.

Phishing remains dominant

Last year, though the global fraud and spam landscape spanned several different message categories, phishing remained the top fraud category. In this method, fraudsters employ deceptive messages to manipulate recipients into divulging sensitive information like credentials, payment details, or other personal data.

In 2025, phishing accounted for 49% of all blocked harmful content, implying that nearly half of the blocked messages were phishing attempts. Phishing represented a substantial 50% of pattern-based detected messages and approximately 57% of all URL-detected messages. YoY, the volume of these messages surged 94%, underscoring the persistent threat of phishing attempts.

Gambling traffic, which represents messages tied to unauthorized or policy-restricted betting activity, were another prominent threat, accounting for 26.4% of all blocked content. Notably, gambling constituted about 42% of AI-powered detected blocked messages, slightly less than 30% of all pattern-based detection, and approximately 15% of URL-detected blocked messages.

Regulatory content ranked third, accounting for approximately 9% of all pattern-based blocked messages in 2025 and 10% of AI-based detected blocked messages. These messages are blocked due to violations of local communication rules or compliance requirements.

The rise of SMS-based fraud

In 2025, SMS continued to be the dominant fraudulent communication channel, accounting for a significant 62% of all platform traffic. This underscores the enduring preference for SMS in situations where reliability and speed of delivery are paramount, such as notifications, one-time passwords, transaction alerts, appointment reminders, or any scenario requiring the prompt transmission of a small, time-sensitive piece of information.

Growth in SMS firewall activity was significant in Asia-Pacific (APAC), and Africa. In 2025, blocked SMS traffic increased 14% in Africa, while APAC experienced even more substantial growth, with blocked traffic increasing by 86% YoY.

Further reflecting the growth of SMS-based fraudulent attempts, the Anam Protect SMS firewall recorded a notable 27% YoY increase in SMS traffic blocked at the operator network level. Tactics in this category include SMS grey-routing, artificially inflated one-time password (OTP) traffic, as well as spam and phishing campaigns.

Anam Protect is a telecom network security solution by Infobip designed for mobile network operators (MNOs). Unlike enterprise security products that operate at the application layer, Anam Protect operates within the operator’s own network infrastructure, providing telecoms with direct control over the application-to-person (A2P) SMS traffic flowing through their systems.

North America and Europe witnessed the most dramatic increases

Among the regions covered by Infobip’s platform data, North America experienced the strongest YoY increase in blocked messages, recording a 224% increase. In the region, phishing attacks accounted for 55% of blocked messages, while regulatory violations accounted for 11%. These figures are significantly higher compared to other regions.

Europe ranked second, with blocked messages in the region growing by 119% in 2025. Europe’s threat profile is characterized by distinctive features, with gambling content leading the way, accounting for 33% of all blocked messages and increasing 352% YoY. This surge indicates a structural expansion of gambling spam operations into European markets, aligning with the growth of both regulated and unregulated online gambling across the continent.

After Europe, Africa followed, with a 118% YoY increase in blocked messages in 2025. Phishing attacks accounted for 15% of blocked traffic, while gambling content accounted for 11%.

Financial services organizations face sophisticated fraud vectors

Looking at sectoral trends, the report shows that fraud exposure varies significantly across different verticals. In the banking, financial services, and insurance industry, fraud exposure spans various stages, including onboarding identity verification, account login, payment authorization, loan confirmation, and insurance claims processing.

Organizations in finance also face the most sophisticated attack vectors because they are high-value targets. According to the Nasdaq Global Financial Crime Report, fraud scams and bank fraud schemes totaled US$485.6 billion in losses in 2023 globally, underscoring the massive economic incentive for fraudsters and criminals.

These organizations are also seeing an expansion of their attack surface through rapid digital modernization. Across all regions, the banking, financial services, and insurance industry ranked as the largest consumers of network APIs, with fintech and finance companies accounting for 77% of all know-your-customer (KYC) APIs.

These interactions validate user-provided identity details against operator subscriber records, allowing businesses to confirm a customer’s identity at the point of onboarding or application without manual document checks.

Featured image: Edited by Fintech News Singapore, based on image by thanyakij-12 via Magnific