AI Is Splintering the Ransomware Market Into Smaller, Faster Crews
- Gate
- Machine Learning
- Departs
- Crew
- Matt Burgess, Lily Hay Newman
- Passengers
- 344
Claimed ransomware attacks have spiked as AI lowers the barrier for lone operators. Fully automated attacks haven't appeared yet, but the trend is moving fast.

The numbers on ransomware are messy, and that's part of the problem. Victims don't always disclose incidents, and attackers have every reason to inflate their claims. Still, the researchers who track and de-duplicate leak-site postings agree on the direction: claimed incidents have jumped over the past few months. Some estimates put alleged attacks posted to leak sites at nearly 3,000 over the summer, a rise of 20 to 40 percent since spring. The more important point is that the baseline itself has stayed elevated for years. This isn't a seasonal blip.
International growth is the new part
Cynthia Kaiser, a former FBI cyber deputy director who now leads research at the anti-ransomware firm Halcyon, says most of the recent increase came from attacks outside the United States. U.S. volume stayed steady, though still high. "Internationally we're seeing a lot more attacks," she says. "That to me is the really new part, and I do think it is in part fueled by AI."
The mechanism she points to is familiar from scams generally. Ransomware gangs often get in through phishing and social engineering, so trends in fraud tend to show up in ransomware numbers later. Kaiser cites Japan, where successful scamming has spiked sharply in recent months. Language has been a long-standing brake on foreign attackers targeting Japanese speakers, and LLMs are removing it. Convincing Japanese-language content is now cheap to produce at scale.
AI helps the crime, not yet the crew
Here's the thing: fully AI-driven ransomware operations haven't surfaced. Part of the reason is that the existing business model already works well enough that attackers have little pressure to rebuild it around automation. What researchers do see is AI slotting into specific steps of the attack chain. Allan Liska, a longtime ransomware researcher and threat intelligence analyst at Recorded Future, says groups are using it to write better phishing emails and to support the negotiation process. The most successful groups, he says, still run largely on human operators.
That said, Liska adds a caveat that deserves attention. Groups are experimenting heavily with these tools, and that experimentation could shift the picture quickly. The gap between "AI helps with the email" and "AI runs the intrusion" is narrower than it looks on paper, and it's the kind of gap that closes in a few product cycles, not a few years.
Fragmentation and the lone-wolf wave
A handful of groups, including Gentlemen, Qilin, and Akira, have emerged as the most prominent and stable operations recently. Beneath them, though, the ecosystem is splintering. Halcyon's analysis counts around 111 ransomware groups that made attack claims between July and September this year. Many are new, and some have moved fast. One newcomer, Global Secret Group, claimed 48 victims, with more than half listed on the same day, according to ReliaQuest's analysis. Another, called Storm, posted 71 victims in just 20 days. Not every claim will hold up, but the claims alone create uncertainty for businesses trying to assess their exposure.
Brandon Tirado, director of threat research at ReliaQuest, reads the surge as a structural shift. "That growth has all happened because, in my eyes, this acceleration that AI has provided to these threat actors, essentially allowing for them to go out and start their own groups," he says. Rather than renting infrastructure from established ransomware-as-a-service providers, individuals are launching their own operations. Every researcher who spoke to the source reported seeing more of these independent operators.
Tirado's team traced one of these new groups to an inexperienced teenager who, by his account, had made more than $800,000 in under two weeks. That single case says a lot about how low the cost of entry has become. He declined to name the teen publicly.
Will the upstarts last?
Kaiser isn't convinced these independent crews will endure. Many may surge and then burn out, she suspects, because they lack the institutional knowledge and resources that established groups have built over years. She also notes that AI is giving these actors more reach: help with tricking targets, finding and exploiting vulnerabilities, checking their own code, and even taking over other groups' remote-access tools to keep a foothold if one path gets blocked.
The operating model is changing too. Established gangs typically run double extortion, encrypting a victim's systems while also stealing data and threatening to publish it. Smaller independent operators may focus purely on data extortion, which is simpler to run. Liska's read on the numbers follows from that. The count of attacks Recorded Future has seen this year already exceeds the full 2025 total, he says, yet the number involving encryption may actually drop year over year.
That combination is worth watching. A higher volume of lower-complexity extortion, run by more operators with less experience, could generate more victims without the technical sophistication that defenders have spent years preparing for. Whether the surge fades or hardens into a durable layer of the threat landscape will depend on how many of these crews outlast their first few weeks.