Dutch cybercrime-as-a-service now fuels 2.5 million victims — and AI is the accelerant

The Netherlands recorded an estimated 2.5 million victims of cybercrime or online fraud in 2025, according to the biannual Cybercrimebeeld Nederland report published this week by the Dutch police and Public Prosecution Service (OM). The headline number is less surprising than what’s driving it: a mature cybercrime-as-a-service economy that lets anyone with cash — and minimal technical skill — assemble complex attacks from off-the-shelf tools, stolen data and rented infrastructure.

“Criminals combine data from earlier hacks with newly stolen data,” said Christa de Pagter, the national public prosecutor for cybercrime. Offenders then use that data to impersonate banks or crypto platforms, pressuring victims into acting fast. The report warns that the attack chain is now modular: specialised groups worldwide collaborate temporarily, then dissolve, making it hard for law enforcement to follow the money or the code. Source: DutchNews.nl

AI makes everything faster — and harder to spot

Artificial intelligence hasn't changed the fundamentals of phishing or malware, but it has dramatically increased speed, scale and believability. According to the OM, criminals use AI to generate phishing messages, develop malware and scan for vulnerabilities. The same technology that powers legitimate code assistants can now produce grammatically perfect scam emails in seconds, stripping away the telltale spelling errors that once gave frauds away. Source: NL Times

A separate study by ABN Amro, released on Friday, underscores the problem: one in seven young adults (18–30) fell victim to online fraud in 2025. That’s higher than the rate for the over-55s, despite young adults rating themselves as more tech-savvy. “The more strongly you believe you can always spot AI, the greater the chance that it will catch you out,” said Marco Hendricks, manager of ABN Amro’s Fraud Helpdesk. Source: NL Times

State actors and teenagers share the same toolkit

The report also highlights a blurring line between criminal gangs and state actors. Countries engaged in hybrid warfare — including Russia — are increasingly using criminal proxies, supplying them with money and infrastructure. In May, Dutch investigators arrested two men over servers linked to Russian cyberattacks; the report also mentions two 17-year-olds detained last year for mapping Wi-Fi networks around government buildings in The Hague.

“These developments underscore the need for a broad approach in which public and private parties cooperate nationally and internationally,” the OM said. The challenge for engineers and security teams is that the same modular, service-based model that powers modern software development now powers cybercrime. Defenders can no longer assume that an attacker needs deep expertise — they just need a credit card.

What this means for builders

For the Tech Desk audience, the takeaway is concrete: the threat model has shifted from lone hackers to a marketplace. If you’re building a product that handles payments, logins or personal data, you should assume that attackers can buy zero-day-like capabilities as a service. The Dutch authorities’ report is a reminder that security design must account for adversaries who can assemble bespoke attack chains faster than most teams can patch.

ABN Amro has responded with a campaign called “Just Begin Safe,” aimed at helping young adults verify online stores and recognise AI-generated scams. But for engineers, the real fix is systemic: build systems that assume the user will make a mistake, and limit the damage when they do.