/
랜섬웨어 격리

The SafePay Ransomware Group Exposes Cybersecurity's Achilles' Heel

A firewall flaw and a misconfigured admin account let SafePay into one global company's network. Within hours, the attackers had domain administrator access. A week later, more than 60 servers were encrypted.

Call it cybersecurity's Achilles' heel. The Greek warrior was nearly invincible in battle, yet one weak spot brought him down. Companies today spend heavily on firewalls, VPNs, and endpoint security to keep attackers out. SafePay is showing what happens when those defenses become the way in.

Ancient Greek vase painting of Achilles in combat during the Trojan War
Ancient Greek polychromatic pottery painting (dating c. 300 BC) of Achilles during the Trojan War.

The ransomware group has hit at least 600 victims worldwide since August 2024.

In October 2026, the FBI, CISA, and international partners issued a joint advisory on how it works. SafePay exploits flaws in security infrastructure, moves through networks with legitimate admin tools, and disables defenses before it encrypts files.

SafePay is worth studying for more than its victim count. Its attacks show that prevention will fail somewhere, and what happens next decides the damage. Teams that can see how their systems connect, and that use segmentation to limit where an intruder can go, can contain a breach before it spreads.

SafePay: new name, old playbook

SafePay emerged in August 2024 and quickly became one of the most active ransomware groups in 2026. Unlike many ransomware-as-a-service groups, SafePay reportedly works as a closed team. Its operators run their own attacks and extortion campaigns instead of relying on affiliates.

But the people behind SafePay may not be newcomers.

Huntress researchers found many similarities between SafePay's ransomware and leaked LockBit Black code. Others see Conti's methods in SafePay's attacks and suspect the group includes former Conti members.

Those ties remain unconfirmed. Still, SafePay may be using tools and tactics that other gangs have used for years.

"It's believed that this group uses similar tactics," said Michael Adjei, Illumio global solutions architecture director. "It's a modification and combination of some of these ransomware groups."

One weak firewall, 60 encrypted servers

In 2025, incident response firm Sygnia investigated a SafePay attack on a global company it didn't name. Its findings, published in March 2026, trace how one gap at the edge became access across the network.

The attackers came in through the company's SSL VPN, using a FortiGate firewall flaw and a misconfigured admin account. Within hours, they had domain administrator access.

From there they moved through the network, going after identity systems, file servers, and backups. They used trusted admin tools, so their activity looked like routine IT work.

The firewall did stop one thing: an attempt to send stolen data out over FTP. So the attackers switched to Microsoft OneDrive and moved the files through a cloud service the company already trusted.

After seven days of data theft, they deployed the ransomware. It encrypted more than 60 servers, including backup, virtualization, and identity systems.

The company restricted network traffic and shut down servers, which held the damage to about 60 systems. By then, the attackers had already reached the systems that mattered most.

How SafePay turns trusted defenses into weaknesses

That case fits a pattern. The joint government advisory describes how the group uses stolen credentials, brute-force attacks, and known flaws in internet-facing systems, including VPNs, firewalls, and backup software.

"They target the security infrastructure itself as a means of getting initial access," Adjei said.

"Those devices are supposed to stop things getting in," he said. "Once attackers bypass them, the next best thing is strong internal segmentation or visibility."

Blending in with legitimate tools

Once inside, SafePay uses tools that make attacks look like routine IT work.

In its investigation of two SafePay intrusions, Huntress found several examples:

  • Attackers used valid credentials and Remote Desktop Protocol (RDP) to gain access.
  • They used ShareFinder to locate network shares and find files on connected systems.
  • In one case, they disabled Windows Defender after it blocked ShareFinder.
  • They used WinRAR to bundle files before deploying ransomware.

These are living-off-the-land techniques. Attackers do their work with tools the company already runs and trusts.

"Once they're in, they blend into the existing environment by living off the land," Adjei said.

The bigger risk is how far attackers can move once they hold stolen credentials or privileged access.

When attackers disable security defenses

Past the perimeter, SafePay goes after the next layer. The advisory says the group typically disables security features such as Windows Defender.

Adjei said the same risk applies to endpoint detection and response tools.

"If you find a way of shutting it down or bypassing it without raising an alarm then you almost have free rein," he said.

Detection still matters. But if attackers evade or disable endpoint security, companies need other ways to stop the spread.

That's where segmentation comes in.

"With segmentation, you don't need to detect an attack first," Adjei said. "You define what's allowed and what isn't."

Segmentation controls which systems can talk to each other. Those rules still apply when security tools miss an attack. Even with stolen credentials or admin access to one machine, an attacker can reach only what policy allows.

SafePay goes after the backups too

The joint advisory says SafePay halts Volume Shadow Copy Service, the Windows feature that keeps restore snapshots. It also resets network-attached storage that holds backup data to factory settings.

Attackers encrypt production systems and destroy backups to make recovery harder, which puts more pressure on victims to pay.

"When we talk about segmentation, that should especially include backup systems or backup storage," Adjei said. "Backups are your last line of defense and recovery."

Teams can cut this risk by limiting admin access, isolating backup storage, and blocking network links that backups don't need. Immutable and offline backups add protection. Teams should also test how well they can restore critical operations.

"If the attacker can get your backups, you don't have backups," he said.

Protect your Achilles' heel

The fix starts with the basics. "The first piece is just standard cyber hygiene," Adjei said. That means patching known flaws, securing credentials, and limiting privileged access.

But prevention alone won't stop every attack.

"That is not in itself a panacea," he said. "You also need to combine that with segmentation, especially of your minimum viable operating systems."

"You have to assume that an attacker will eventually get in," he said. "The question is how far they can go. Segmentation limits their reach, so one compromised system doesn't put the entire business at risk."

Achilles never got the chance to guard his heel. Security teams do. Every perimeter has a weak point somewhere. With segmentation, the attacker who finds it gets one system, and the rest of the network stays out of reach.

Contain ransomware with Illumio Segmentation

When perimeter defenses fail, what counts is how far an attacker can move.

Illumio Segmentation helps contain ransomware by blocking the connections your systems don't need, so a breach in one place stays there.

Illumio Segmentation 체험하기

관련 문서

지금 일루미오 인사이트 체험하기

AI 기반 관찰 가능성을 통해 위협을 더 빠르게 탐지, 이해, 차단하는 방법을 알아보세요.