HomeBlogThreat Watch

3 New Malware & Virus Threats Hitting Business PCs in 2026

Three glowing red virus particles circling a blue security dome that shields a digital city, representing 2026 malware threats to corporate computers

The new computer virus names in recent CISA and FBI advisories are mostly ransomware and password stealers: Gunra, Medusa, Akira and Interlock ransomware, and the LummaC2 infostealer, often delivered through fake CAPTCHA pages known as ClickFix. Together with the first malware that asks AI models for help, such as PROMPTSTEAL, they make up the three threats hitting business PCs in 2026: ransomware that steals data before it encrypts, credential-stealing infostealers, and AI-assisted phishing. Here is each name, how it gets in and how to stop it.

"Virus" is still the word most people use, but the threats doing the damage now are malware families run by criminal groups, often rented to affiliates as a service, that chain stolen passwords, legitimate remote tools and social engineering. Few of them need anyone to open an obviously suspicious file; most arrive looking like an invoice, a browser update or a security check.

New computer virus names to know in 2026

These are the malware names in recent joint advisories from CISA and the FBI, plus Google's threat intelligence reporting on malware that uses AI, as of October 2026:

NameTypeHow it gets inSource
GunraRansomware-as-a-service based on leaked Conti code; first seen in April 2025Known flaws and default credentials in internet-facing firewalls and VPN gatewaysCISA advisory AA26-222A, August 10, 2026
MedusaRansomware-as-a-service with double extortion; over 500 victims as of April 2026Phishing for passwords, and unpatched software including a BeyondTrust flaw disclosed in February 2026CISA advisory AA25-071A, updated August 18, 2026
AkiraRansomware that mainly targets small and medium-sized businessesVPNs without multifactor authentication, SonicWall and Veeam vulnerabilities, stolen VPN credentialsCISA advisory AA24-109A, updated November 13, 2025
InterlockRansomware with double extortion; first seen in September 2024Compromised legitimate websites, fake browser and security updates, ClickFix pagesCISA advisory AA25-203A, July 22, 2025
LummaC2 (Lumma Stealer)Infostealer sold to other criminals since 2022Spearphishing links and attachments, fake CAPTCHA pages, fake versions of popular softwareCISA advisory AA25-141B, May 21, 2025
ClickFixA delivery trick rather than one malware familyA fake CAPTCHA tells the visitor to paste a command into the Windows Run box, which launches PowerShellDescribed in the LummaC2 and Interlock advisories
PROMPTSTEAL (also reported as LAMEHUG)Data-stealing malware that asks an AI model to write its Windows commandsUsed by the Russian state-backed group APT28 against UkraineGoogle Threat Intelligence Group, November 5, 2025
PROMPTFLUXExperimental malware that asks Google's Gemini to rewrite its own codeGoogle says it was still in testing and could not yet compromise a deviceGoogle Threat Intelligence Group, November 5, 2025

AI-Adaptive Ransomware: the Most Disruptive Endpoint Threat

Ransomware remains the most damaging malware on corporate computers. Verizon's 2026 Data Breach Investigations Report, covering incidents from November 2024 through October 2025, found ransomware in 48% of breaches, and found that 31% of breaches now start with an exploited software vulnerability, ahead of stolen passwords as the top way in.

Modern attacks rarely encrypt files the moment they arrive. Gunra, Medusa, Akira and Interlock all use double extortion: the attackers steal data first, then encrypt systems and threaten to publish what they took, which is why good backups alone no longer end the crisis. The AI part is real but early. Google's threat intelligence team has documented malware that asks AI models for commands in live attacks, and experimental code that rewrites itself, but the ransomware in recent advisories still gets in mostly through stolen credentials, unpatched VPNs and remote access tools.

Corporate endpoints are central to these attacks because employee workstations provide access to identity systems, cloud services and internal networks. Once administrative or VPN credentials are captured from a compromised computer, attackers can escalate access far beyond the original machine. Our guide to preventing ransomware in a small business covers the defenses in order.

Credential-Stealing Infostealers: Silent Gateways

A second major threat to corporate computers is credential-stealing infostealer malware. CISA lists LummaC2 stealing personal and financial credentials, browser data, cryptocurrency wallets and multifactor authentication details, and notes that it is traded in volume: more than 21,000 market listings of LummaC2 logs appeared from April through June 2024.

Infostealers arrive through phishing emails, fake invoices, fake CAPTCHA pages and fake versions of popular software. Because they work through the browser and ordinary encrypted web traffic, their activity can look like normal use, and CISA recommends application allowlisting partly because antivirus may miss obfuscated programs.

The risk is significant. Stolen sign-in data from one laptop can open other people's email, cloud apps and file storage without any further malware, and stolen session tokens can get past multifactor authentication, as our guide to session hijacking explains. Infostealers are often the first step of a ransomware attack: CISA notes that researchers have seen Interlock infections use the Lumma and Berserk stealers to harvest credentials.

AI-Generated Phishing Malware: Delivery Becomes the Weapon

Phishing remains a primary delivery route in the advisories above: Medusa operators use it to steal credentials, and LummaC2 spreads through spearphishing links and attachments. Generative AI makes the lures more convincing, because messages now arrive in fluent language that matches the tone of internal mail. Follow-up messages come through Microsoft Teams, text or voice calls, and attachments may be HTML, SVG or links to cloud-hosted documents rather than obvious programs.

The newest twist skips attachments entirely. ClickFix pages show a fake CAPTCHA and tell the visitor to paste a command into the Windows Run box, and CISA describes that route in both the LummaC2 and Interlock advisories. Because the user runs the command, email filters never see it, which is why knowing the ClickFix trick and ongoing security awareness training matter as much as filtering.

Why These Threats Matter for Your Business

All three threat types rely on corporate computers and user behavior. Endpoints now hold saved passwords, cloud sessions and access to every app a person uses, so one compromised laptop reaches much further than it once did.

Antivirus software alone is no longer enough: EDR and managed detection go further than antivirus. Layered defenses, such as endpoint detection and response (EDR) watched by people, phishing-resistant multifactor authentication and continuous user awareness, close the gaps that today's attack chains use. Our managed detection and response service covers the monitoring side, and our ransomware protection work covers backups and recovery.

How do you protect business PCs from these threats?

The same mitigations repeat across the CISA advisories above. In order of impact for a small business:

  1. Patch internet-facing devices first. Firewalls, VPN gateways and remote access tools are how Gunra and Akira get in, so apply fixes for known exploited vulnerabilities quickly.
  2. Require phishing-resistant multifactor authentication for VPNs, email and admin accounts, and remove default and unused accounts.
  3. Keep offline, immutable backups and test restores, so encryption does not end the business.
  4. Run EDR on every computer and server, with someone watching the alerts around the clock.
  5. Limit admin rights and scripting. Separate admin accounts from daily accounts, and restrict command-line and scripting tools for staff who do not need them, which also blunts ClickFix.
  6. Segment the network so one infected PC cannot reach every server.
  7. Train staff on current lures, including fake CAPTCHAs, fake updates and invoice scams.

What should you do if a PC has a virus?

  1. Disconnect it from the network by unplugging the cable or turning off Wi-Fi, and do not wipe or reinstall it until it has been examined.
  2. Call your IT provider or security team before trying fixes yourself.
  3. Reset the user's passwords from a clean device and sign them out of every active session.
  4. Check your backups before anyone restores anything.
  5. Report it. CISA and the FBI ask victims to report incidents to the FBI's Internet Crime Complaint Center (IC3), a local FBI field office or CISA, and they do not encourage paying a ransom. Your insurer, and in regulated industries your regulator, may need notice too.

The Bottom Line

The most significant virus threats to corporate computers in 2026 are not isolated pieces of malicious code but coordinated attack chains that use employee endpoints as the way into the business. Ransomware that steals before it encrypts, credential-stealing infostealers and AI-assisted phishing are the patterns behind the names above, and the names change faster than the defenses that stop them. Businesses that patch their edge devices, protect sign-ins and watch their endpoints are far better placed to limit the damage.

Quick summary

Threat typeWhy it is dangerous
AI-adaptive ransomwareSteals data before encrypting, so backups alone do not end the crisis
InfostealersQuiet, uses the browser and normal web traffic, and opens the door to a full breach
AI-assisted phishingFluent lures, fake CAPTCHAs and links to legitimate cloud services that get past filters and people

Infostealers are not the only thing reading the browser; our guide to browser extension security risks shows how approved add-ons see every page after login, and how allowlisting stops that.

If the acronym is new to you, our glossary explains what endpoint detection and response is: software that records activity on each computer, spots attacker behavior and isolates the device.

Frequently asked questions

What are the newest computer virus names?

The names in recent CISA and FBI advisories are Gunra ransomware (August 2026), Medusa ransomware (updated August 2026), Akira and Interlock ransomware, and the LummaC2 infostealer, often delivered through ClickFix fake CAPTCHA pages. Google has also reported the first malware that calls AI models, PROMPTSTEAL, and an experimental one, PROMPTFLUX.

Are antivirus programs still effective against modern threats?

Alone, no. Modern attacks combine malware with stolen credentials, legitimate tools and social engineering, so signature-based antivirus routinely misses them. Organizations need layered defenses, such as endpoint detection and response (EDR), multifactor authentication and continuous user awareness, to close the gaps that antivirus leaves open.

What are the most common malware threats to businesses in 2026?

The three biggest threats to corporate computers in 2026 are ransomware that steals data before encrypting it, credential-stealing infostealers, and AI-assisted phishing. Verizon's 2026 report found ransomware in 48% of breaches. All three are coordinated attack chains rather than isolated pieces of code, and they use employee endpoints as the way into identity systems, cloud services and internal networks.

How do infostealers steal passwords without triggering antivirus?

Infostealers harvest saved browser passwords, cloud sign-in tokens and session data while working through legitimate browsers and encrypted web traffic. Because that activity looks like normal use, endpoint protection tools can struggle to flag it. The malware arrives through phishing emails, fake invoices, fake CAPTCHA pages or ordinary-looking downloads, and quietly captures credentials.

What does managed detection and response include, and who handles escalations?

Managed detection and response pairs EDR software on every device with people who watch the alerts, investigate them and isolate an infected machine. At NetSys, monitoring and emergency service run 24/7, and our help desk is staffed seven days a week from 4 a.m. to 11 p.m. Eastern. Whoever you use, the agreement should say who contains a threat, who calls you, and what happens after hours.

Sources and further reading

Cyber Security

Discuss cyber security for your business.

Tell us about your current systems, the result you need and your timeline. We will discuss the work, responsibilities and pricing before you decide on an engagement.