Researchers link AI tools to a rise in ransomware from small and inexperienced groups
Fully automated ransomware has not arrived. Researchers say AI is helping lone operators, including one teenager, launch attacks that once needed a skilled team.
Claimed ransomware attacks have risen sharply in recent months, and researchers who track the groups say AI tools are part of the reason. Machines are not running the attacks from start to finish. What has changed is who can launch one: people with little experience are using language models for the parts of the job that used to require skill or a team.
The report comes from Wired's Kernel Panic newsletter by Lily Hay Newman and Matt Burgess. They spoke to researchers at Halcyon, Recorded Future and ReliaQuest.
A rise in claimed attacks
Exact counts are hard to get. Victims do not always disclose incidents, and attackers inflate their claims. Researchers who track leak sites, where groups post alleged victims, and remove duplicate claims still agree that claims have spiked. Some put the number of alleged attacks listed over the summer at nearly 3,000. By some estimates that is an increase of 20 to 40 percent since the spring. Researchers stress that volume has stayed high for years, whatever the short-term swings.
Cynthia Kaiser, a former FBI cyber deputy director who now leads research at the anti-ransomware platform Halcyon, says most of the recent increase came from attacks outside the United States. Numbers in the US stayed "a little more steady, but still high." She calls the international growth "the really new part" and says it is "in part fueled by AI."
Allan Liska, a threat intelligence analyst at Recorded Future, says the firm's count of attacks this year "already dwarfs the total number for 2025."
Where AI shows up in the attack chain
Fully AI-powered ransomware has not appeared. Wired suggests one reason may be that the existing criminal business model already works well. What researchers do see is AI used for routine tasks.
"We are seeing AI relied on more heavily in certain aspects of the ransomware attack chain, such as creating better phishing emails and in the negotiation process, but the 'successful' ransomware groups are still largely human operated," Liska says. Because these groups are experimenting heavily, he adds, "this could change going forward, and change quickly."
Kaiser points to language as one example. Ransomware often starts with the same phishing and social engineering used in ordinary scams. In Japan, successful scams have risen sharply in recent months. AI now lets foreign attackers write consistently convincing Japanese for the first time. A language barrier that once kept many attackers out has weakened.
She lists other tasks AI now helps with: tricking targets, exploiting vulnerabilities and checking code. Some attackers also use it to break into other groups' remote-access tools, which gives them a backup way into a victim's network if one route is blocked.
Smaller groups, and one teenager
Three operations, the Gentlemen, Qilin and Akira, have become the most prominent and stable in recent months. Around them, the field is splitting into many smaller groups. Halcyon's analysis counts about 111 ransomware groups that claimed attacks between July and September.
New names appear fast. According to ReliaQuest, a group calling itself Global Secret Group claimed 48 victims, more than half of them listed on a single day. Another, Storm, listed 71 victims in 20 days. Not all of these claims may be real. Even false ones create uncertainty for the businesses named.
Brandon Tirado, director of threat research at ReliaQuest, attributes the growth to "this acceleration that AI has provided to these threat actors, essentially allowing for them to go out and start their own groups." Every researcher Wired spoke to described more lone operators working outside the usual ransomware-as-a-service model, where attackers depend on an established provider. Tirado says his team identified the person behind one new group: an inexperienced teenager who made more than $800,000 in under two weeks. He declined to name the teenager publicly.
Kaiser is unsure how long these newcomers will last. She expects many to surge and then burn out, because they lack the institutional knowledge and resources of established groups.
The solo operators may also be changing what an attack looks like. Established groups usually run "double extortion": they encrypt a victim's systems, steal data and threaten to publish it. Independent actors may skip encryption and rely on data extortion alone, which is simpler to run. Liska says total attacks are up, but "the number of attacks involving encryption may actually decrease year over year."
What it means for work
The tasks AI is taking over here used to need a skilled team or a rented service: writing believable phishing in a foreign language, handling ransom negotiations, checking code and finding a way in. That lowers the cost of entry and moves capability from organised groups to individuals. Defenders face more claims to triage, some of them possibly false. If more attacks are pure data theft without encryption, more of their work may shift from restoring systems to protecting and monitoring data. People who read email in languages that once offered some protection, such as Japanese, lose a warning sign they relied on. Two questions remain open: whether the new operators last, and whether the human-run groups that dominate today move further toward automation, which Liska warns could happen quickly.
Sources
This article was researched, written and illustrated by AI, and published without human editing. Spotted an error? Let us know.
More from māchine
All storiesGoogle has added computer use to the Gemini app for macOS. Paying US subscribers can hand it multi-step tasks, decide which apps it may touch and approve its key actions.
Yale researcher Julián Posada argues that platforms recruiting doctors, lawyers and PhDs to train AI pay well at first, then cut rates and drop workers, as they did in Venezuela and Kenya.
Microsoft offered voluntary exits to about a quarter of its workforce in October. Its CEO says the reason is AI, as a reorganisation of work around managing agents, not as a replacement for people.