Visa has released its mid-year 2026 Biannual Threats Report, and the takeaway is clear: while network security is tighter than ever, fraudsters have pivoted. Instead of trying to crack sophisticated firewalls, cybercriminals are using artificial intelligence and social engineering to exploit human trust, making scams the fastest-growing source of consumer financial harm.
Between July and December 2025, Visa identified nearly $1 billion in scam-related activity, cementing scams as the single largest category of consumer payment fraud. Rather than breaching technical systems, modern scammers simply impersonate trusted brands, manufacture artificial urgency, and trick victims into willingly authorizing transactions.
Fraud trends at a glance
The landscape is shifting rapidly. While defensive technology is successfully blocking traditional entry points, criminals are finding success through volume and social manipulation.
| Metric / Threat Vector | Performance (H2 2025 vs. H2 2024) | Implications |
| Scam-Related Activity | ~$1 Billion (Total Volume) | Now the #1 threat to consumer payment security. |
| Device Token Fraud | ↘️ 9.6% Decrease | Confirms that stronger network-level authentication is working. |
| Ransomware Activity | ↗️ 26% Increase | Attack volumes are up, but corporate resilience is improving. |
| Ransom Payment Rate | ↘️ 23% (Record Low) | Victims are refusing to pay, knowing data leaks happen anyway. |
Trends reshaping Global payment security
1. Stronger security is forcing fraud to migrate
Network protections are doing their job. The 9.6% drop in token-related fraud proves that advanced authentication methods are highly effective. However, because the front door is locked, criminals are shifting their focus to manipulating the person holding the keys.
2. Scams are accelerating
Social engineering has officially overtaken technical system breaches as the preferred method of attack. By mimicking legitimate institutions, scammers trick consumers into bypass-proof actions, because the consumers are the ones clicking approve.
3. AI is a double-edged sword
Generative AI has fundamentally altered the playing field for both sides:
- The Bad: Fraudsters use AI to scale highly convincing, hyper-personalized scams with zero technical skill required.
- The Good: Defenders are aggressively deploying AI to analyze transaction lifecycles and intercept fraudulent patterns before the money leaves the account.
4. Ransomware economics are imploding
While global ransomware incidents surged by 26%, the profitability of these attacks hit an all-time low. Only 23% of victims paid out. Organizations are developing better backup and recovery capabilities, alongside a growing realization that paying a ransom rarely guarantees data privacy.