Cybercriminals exploited artificial intelligence in ransomware attacks suffered by 84% of victims surveyed, according to a new global cybersecurity report, while 44% said the attackers exploited their own AI systems.
The 2026 Ransomware Risk Report from cybersecurity firm Semperis surveyed 1,200 IT and security professionals across nine countries and eight industry sectors, revealing an evolved picture of the cybercriminal enterprise, which is increasingly leveraging AI to amplify its technical capabilities and increase pressure on its targets.
But the findings also highlighted the human toll of ransomware. According to the report, more than a third of victims surveyed have faced psychological or physical threats against their employees, and 14% of organizations surveyed even reported that employees or family members had been forced to change jobs, offices, or residences due to sustained ransomware threats.
In total, 69% of ransomware victims surveyed experienced multiple successful attacks in the past year, with 59% of respondents having paid a ransom at least once, and 87% were victims of attacks targeting their identity systems or the digital infrastructure that controls access to their networks.
For Chris Inglis, former national cybersecurity director at the White House under former President Joe Biden, artificial intelligence represents a turning point in the capabilities conducive to cybercriminal exploits – and he warns that one particularly consequential development is already underway: autonomous ransomware attacks.
“AI actually increases the capability and scope and scale of attackers by allowing them to understand what the weaknesses might be in your system or perhaps increase the sophistication of their approach to you if they’re trying to get you to do something you shouldn’t do, it gives them an advantage, lets them in,” Inglis said in an interview about the report.
The cybersecurity expert and Semperis board member has already spent three decades watching technology evolve within the National Security Agency, observing that attacks required far less direct human involvement – no longer just a theoretical possibility.
“AI is an inflection point in their ability to endanger the same things,” Inglis said. He later added: “I think we are already seeing autonomous attacks. Standalone ransomware attacks are included in the mix. »
Inglis described how someone without the technical expertise to penetrate an organization’s networks could turn to AI to write the necessary code, thereby instructing an AI system to target a company and hold its data hostage without a deep understanding of how the digital infrastructure works.
“We’ve democratized AI to the point where even low-level models can allow this to happen,” Inglis said, noting that the technology increases the number of criminals capable of launching more technical attacks, while inviting experienced cybercriminals to penetrate networks more effectively and on a much larger scale.
AI is a game changer for ransomware
For years, criminals have relied on a familiar pattern: exploit vulnerabilities, penetrate corporate networks, encrypt data, and demand ransom payment in exchange for restoring access or withholding sensitive information. Now, in more than eight out of ten attacks, AI has been integrated into this process.
AI has also allowed attackers to quickly identify weaknesses, tailor their social engineering programs to specific employees, automate an intrusion and readjust their tactics when defenders try to respond, according to Inglis.
“What hasn’t changed are the things that people or organizations owned that have value and that they would pay money or time or energy to protect,” Inglis said. “And these things always receive a lot of attention from criminals. They want to use every tool possible to try to put people in danger so they can get you to pay them money, sometimes time, but really it’s about money.”
By lowering technical barriers, AI risks allowing more actors to attempt intrusions that previously required a wide range of resources and more specialized training or know-how.
Tomer Bar, associate vice president of security research at Semperis, warns that AI could allow ransomware developers to produce more sophisticated malicious code, reduce development time and make investigations more difficult.
“We are starting to see more attacks with faster development, and attribution of these attacks will be more difficult for law enforcement,” Bar says in the report.
“We will also be less likely to see errors in attack encryption algorithms because attackers can test their code with (large language models.)”
When a company’s AI becomes a weapon against it
One of the report’s most striking revelations is that 44% of victims surveyed were victims of ransomware attacks exploiting their own AI agents or systems, a finding that raises an uncomfortable question for businesses and government agencies racing to integrate the technology into their networks.
Courtney Guss, associate vice president of crisis management at Semperis, warns that autonomous AI agents can create security risks when they request additional permissions or interact with sensitive enterprise systems.
“If your business adopts AI agents, you should consider that at some point these agents will attempt to compromise your identity systems, even if this action is not intended to be malicious,” Guss writes in the report.
“That’s exactly how they work. AI agents generally try to accomplish their assigned tasks by any means possible.” Guss continued. “One way for them to gain more ability to operate in the environment is to gain permissions on other systems. The quickest path to gaining these permissions is to compromise your identity system, which effectively puts the agent into ‘God mode,’ allowing it to grant itself the permissions it needs to accomplish its task. Unfortunately, the agent then becomes a ticking time bomb.”
Unlike conventional software, some AI agents possess the ability to use multiple applications at once and make informed decisions on behalf of their users.
If an attacker manipulates just one of these agents (or compromises the credentials and permissions they rely on), they gain another way to penetrate deeper into an organization’s systems.
Inglis, however, cautions against assuming that AI will necessarily make an organization more vulnerable to outside intruders.
“A lot of attackers are already making what we call living off the land: They’re setting up shop in your digital infrastructure, in your property or your digital property, and they’re using the tools that you have, the information or the data set,” Inglis said. “To blend in, to navigate the landscape, they’re trying not to stand out, and they’re doing that by using your tools. If AI is one of your tools, they’ll use them, but the weakness that let them in is probably the same weakness that was there before.”
Inglis says the fundamental vulnerability often lies in a company’s inability to control access to its networks or digital infrastructure. Rather than abandoning AI out of fear, Inglis argues that organizations must harness its defensive potential.
“I would turn it around. Flip the script,” he said. “You need to use AI to counter this threat, and you need to use AI for positive purposes.”
Julia Kimani-Burnham contributed to this report.
Gn bussni
