🔍 SharkLoader pushes Cobalt Strike into memory
Researchers identified SharkLoader, an evasive multi-stage loader used by the StrikeShark cluster to deploy Cobalt Strike Beacon directly in memory. Initial access combines exploitation of internet-facing systems with fake installers, while execution chains rely on DLL side-loading, reflective loading, encrypted stages, and Perfect DLL Hijacking. Reported victims include government, diplomatic, and software targets across Asia, Europe, the Middle East, and Latin America.
The operational value is in low-noise post-compromise access. In-memory execution, signed binary abuse, process spoofing, event log interference, and frequent scheduled-task persistence compress detection time while enabling reconnaissance, credential theft, and Active Directory enumeration.
🛰️ Open sources - closed narratives
@sitreports
Researchers identified SharkLoader, an evasive multi-stage loader used by the StrikeShark cluster to deploy Cobalt Strike Beacon directly in memory. Initial access combines exploitation of internet-facing systems with fake installers, while execution chains rely on DLL side-loading, reflective loading, encrypted stages, and Perfect DLL Hijacking. Reported victims include government, diplomatic, and software targets across Asia, Europe, the Middle East, and Latin America.
The operational value is in low-noise post-compromise access. In-memory execution, signed binary abuse, process spoofing, event log interference, and frequent scheduled-task persistence compress detection time while enabling reconnaissance, credential theft, and Active Directory enumeration.
🛰️ Open sources - closed narratives
@sitreports
🔍 Avalon framework adds ransomware delivery to modular malware stack
A newly documented Avalon malware framework now includes capabilities tied to CrownX ransomware, combining payload delivery, persistence, and post-compromise execution inside a broader modular toolkit. The framework appears positioned to support multiple intrusion stages rather than a single-use payload.
The key significance is consolidation: ransomware functionality is being embedded into flexible malware infrastructure, reducing the gap between initial access and encryption. For defenders, this raises the value of detecting framework behavior early, before operators shift from foothold to disruptive monetization.
🛰️ Open sources - closed narratives
@sitreports
A newly documented Avalon malware framework now includes capabilities tied to CrownX ransomware, combining payload delivery, persistence, and post-compromise execution inside a broader modular toolkit. The framework appears positioned to support multiple intrusion stages rather than a single-use payload.
The key significance is consolidation: ransomware functionality is being embedded into flexible malware infrastructure, reducing the gap between initial access and encryption. For defenders, this raises the value of detecting framework behavior early, before operators shift from foothold to disruptive monetization.
🛰️ Open sources - closed narratives
@sitreports
🔍 DOM Poisoning Expands the AI Agent Attack Surface
Zscaler ThreatLabz documented two campaigns using hidden prompt instructions in website DOM elements to manipulate AI agents. The observed sites combined SEO poisoning, JSON-LD metadata, off-screen CSS text, and typosquatting to feed malicious directives to scrapers and LLM-based tools, including payment instructions and false authority signals. The technique is outlined in DOM poisoning research published on 3 July.
The cases show that AI agents parsing web content can be steered by invisible page elements that do not affect human users. Operationally, this turns search ranking, structured metadata, and retrieval pipelines into a single attack chain against automated troubleshooting, coding, and DeFi-related workflows.
🛰️ Open sources - closed narratives
@sitreports
Zscaler ThreatLabz documented two campaigns using hidden prompt instructions in website DOM elements to manipulate AI agents. The observed sites combined SEO poisoning, JSON-LD metadata, off-screen CSS text, and typosquatting to feed malicious directives to scrapers and LLM-based tools, including payment instructions and false authority signals. The technique is outlined in DOM poisoning research published on 3 July.
The cases show that AI agents parsing web content can be steered by invisible page elements that do not affect human users. Operationally, this turns search ranking, structured metadata, and retrieval pipelines into a single attack chain against automated troubleshooting, coding, and DeFi-related workflows.
🛰️ Open sources - closed narratives
@sitreports
🤖 JadePuffer shows end-to-end ransomware automation
Researchers tracking JadePuffer describe what they assess as the first documented ransomware intrusion run entirely by an LLM agent. The operation exploited CVE-2025-3248 in Langflow, harvested credentials, moved into Alibaba Nacos infrastructure, established persistence, and encrypted 1,342 configuration items using MySQL functions before dropping extortion tables.
The notable detail is not just automation, but adaptation: the agent reportedly adjusted payload logic after errors, retried failed steps with refined parameters, and completed lateral movement and encryption without a visible human operator. That lowers execution barriers while creating a distinct trail of machine-generated code and behavior defenders can baseline.
🛰️ Open sources - closed narratives
@sitreports
Researchers tracking JadePuffer describe what they assess as the first documented ransomware intrusion run entirely by an LLM agent. The operation exploited CVE-2025-3248 in Langflow, harvested credentials, moved into Alibaba Nacos infrastructure, established persistence, and encrypted 1,342 configuration items using MySQL functions before dropping extortion tables.
The notable detail is not just automation, but adaptation: the agent reportedly adjusted payload logic after errors, retried failed steps with refined parameters, and completed lateral movement and encryption without a visible human operator. That lowers execution barriers while creating a distinct trail of machine-generated code and behavior defenders can baseline.
🛰️ Open sources - closed narratives
@sitreports
🔍 North Korean-linked PolinRider campaign pushed 108 malicious dev packages
Researchers tracked 108 malicious npm packages and browser extensions tied to the PolinRider campaign, attributed to North Korean operators. The packages targeted developers and cryptocurrency users, using supply-chain delivery to plant malware through software dependencies and extensions.
The scale matters operationally: 108 separate uploads increase discovery surface and persistence across ecosystems developers trust by default. This is a broad access operation aimed at credential theft and downstream compromise through routine package installation rather than direct intrusion.
🛰️ Open sources - closed narratives
@sitreports
Researchers tracked 108 malicious npm packages and browser extensions tied to the PolinRider campaign, attributed to North Korean operators. The packages targeted developers and cryptocurrency users, using supply-chain delivery to plant malware through software dependencies and extensions.
The scale matters operationally: 108 separate uploads increase discovery surface and persistence across ecosystems developers trust by default. This is a broad access operation aimed at credential theft and downstream compromise through routine package installation rather than direct intrusion.
🛰️ Open sources - closed narratives
@sitreports
🔍 TimbreStealer delivered via trusted updater binaries
A targeted phishing campaign is using ZIP archives hosted on DigitalOcean IPs to deliver TimbreStealer through DLL side-loading with legitimate EdgeUpdate and GoogleUpdater binaries. Victims appear primarily to be companies in Mexico, with lures themed around invoicing and CFDI documents. The payload steals browser, mail, and cloud-synced data, while using RC4-based decryption, PEB/export parsing, and anti-analysis checks.
The tradecraft blends trusted executables with oversized malicious DLLs, geofencing, and runtime-only payload assembly, reducing static detection and complicating reverse engineering. Key indicators include updater-named DLLs around 45–50 MB, ZIP delivery from direct cloud IPs, and suspicious access to browser SQLite files.
🛰️ Open sources - closed narratives
@sitreports
A targeted phishing campaign is using ZIP archives hosted on DigitalOcean IPs to deliver TimbreStealer through DLL side-loading with legitimate EdgeUpdate and GoogleUpdater binaries. Victims appear primarily to be companies in Mexico, with lures themed around invoicing and CFDI documents. The payload steals browser, mail, and cloud-synced data, while using RC4-based decryption, PEB/export parsing, and anti-analysis checks.
The tradecraft blends trusted executables with oversized malicious DLLs, geofencing, and runtime-only payload assembly, reducing static detection and complicating reverse engineering. Key indicators include updater-named DLLs around 45–50 MB, ZIP delivery from direct cloud IPs, and suspicious access to browser SQLite files.
🛰️ Open sources - closed narratives
@sitreports
🔍 CrownX embedded in Avalon hits backup and recovery paths
Researchers tracking Avalon describe a multi-stage framework delivering the CrownX ransomware via phishing, Proton Drive-hosted archives, and a mounted ISO with a fake PDF shortcut. The chain uses MSBuild and in-memory .NET loading, disables ETW and AMSI visibility, manually maps payloads, steals credentials, and later encrypts files while targeting VSS, shadow copies, WinRE, and restore settings.
Operationally, the combination of credential theft, lateral movement, anti-forensics, and recovery disruption in one framework compresses defender response time. Priority target sets included domain controllers, backup platforms, virtualization infrastructure, and other systems critical to restoration.
🛰️ Open sources - closed narratives
@sitreports
Researchers tracking Avalon describe a multi-stage framework delivering the CrownX ransomware via phishing, Proton Drive-hosted archives, and a mounted ISO with a fake PDF shortcut. The chain uses MSBuild and in-memory .NET loading, disables ETW and AMSI visibility, manually maps payloads, steals credentials, and later encrypts files while targeting VSS, shadow copies, WinRE, and restore settings.
Operationally, the combination of credential theft, lateral movement, anti-forensics, and recovery disruption in one framework compresses defender response time. Priority target sets included domain controllers, backup platforms, virtualization infrastructure, and other systems critical to restoration.
🛰️ Open sources - closed narratives
@sitreports
🔍 BusySnake expands Windows infostealer tradecraft
BusySnake is a Python-based Windows infostealer linked in reporting to Armored Likho. It steals browser passwords and cookies, Telegram Desktop session data, clipboard contents, screenshots, OTP-related data, and cryptocurrency wallet files or keys. Delivery uses spear-phishing archives with NSIS droppers or weaponized LNK files, followed by staged loaders, embedded Python 3.12, and PyArmor-protected payloads.
The combination of Telegram session theft, browser credential extraction, and persistent C2 polling gives operators both immediate account access and flexible follow-on tasking. Defenders should treat infections as full credential and session compromise, especially where %APPDATA%-based Python runtimes, abnormal LNK chains, or scheduled VBScript launchers are present.
🛰️ Open sources - closed narratives
@sitreports
BusySnake is a Python-based Windows infostealer linked in reporting to Armored Likho. It steals browser passwords and cookies, Telegram Desktop session data, clipboard contents, screenshots, OTP-related data, and cryptocurrency wallet files or keys. Delivery uses spear-phishing archives with NSIS droppers or weaponized LNK files, followed by staged loaders, embedded Python 3.12, and PyArmor-protected payloads.
The combination of Telegram session theft, browser credential extraction, and persistent C2 polling gives operators both immediate account access and flexible follow-on tasking. Defenders should treat infections as full credential and session compromise, especially where %APPDATA%-based Python runtimes, abnormal LNK chains, or scheduled VBScript launchers are present.
🛰️ Open sources - closed narratives
@sitreports
🔍 Attested TLS flaw cuts into confidential computing trust
New academic work and a published CVE show attested TLS can validate a genuine TEE while still relaying a client session to a different malicious endpoint. The issue, tracked as CVE-2026-33697, affects multiple intra-handshake attestation designs and was identified in production-linked implementations including Meta Private Processing, Edgeless Contrast, and Cocos AI versions 0.4.0-0.8.2.
Operationally, the finding undercuts a core assurance sold by confidential computing: proving the party at the other end of the connection. Standards bodies have acknowledged the relay class, while the research argues the strongest binding to application data may be unreachable in current intra-handshake designs.
🛰️ Open sources - closed narratives
@sitreports
New academic work and a published CVE show attested TLS can validate a genuine TEE while still relaying a client session to a different malicious endpoint. The issue, tracked as CVE-2026-33697, affects multiple intra-handshake attestation designs and was identified in production-linked implementations including Meta Private Processing, Edgeless Contrast, and Cocos AI versions 0.4.0-0.8.2.
Operationally, the finding undercuts a core assurance sold by confidential computing: proving the party at the other end of the connection. Standards bodies have acknowledged the relay class, while the research argues the strongest binding to application data may be unreachable in current intra-handshake designs.
🛰️ Open sources - closed narratives
@sitreports
🔍 U.S. government entity paid $1 million in Kairos extortion case
A U.S. government entity paid $1 million to the Kairos group after a data-theft extortion incident, as detailed in The Hacker News report. The case centers on extortion tied to stolen data rather than disruption or encryption, underscoring the continued effectiveness of pure theft-based pressure operations.
The payment is significant because it shows a government target entering direct financial resolution in a data-leak scenario. Operationally, it highlights that exfiltration alone can generate strategic leverage even without a destructive network impact.
🛰️ Open sources - closed narratives
@sitreports
A U.S. government entity paid $1 million to the Kairos group after a data-theft extortion incident, as detailed in The Hacker News report. The case centers on extortion tied to stolen data rather than disruption or encryption, underscoring the continued effectiveness of pure theft-based pressure operations.
The payment is significant because it shows a government target entering direct financial resolution in a data-leak scenario. Operationally, it highlights that exfiltration alone can generate strategic leverage even without a destructive network impact.
🛰️ Open sources - closed narratives
@sitreports
🔍 Verified X ads used in ClickFix-style macOS malware delivery
A sponsored post from a verified X account redirected users to a fake DynamicLake site and prompted them to paste a copied Terminal command, installing macOS malware including Atomic Stealer variants. The ad was later removed after Jamf alerted X and the account owner.
The case shows how paid placement, verification signals, and lookalike domains can be combined to bypass trust checks while keeping execution user-driven. For defenders, the key indicators are sponsored social lures, clipboard-based Terminal prompts, and brand impersonation tied to fake utility downloads.
🛰️ Open sources - closed narratives
@sitreports
A sponsored post from a verified X account redirected users to a fake DynamicLake site and prompted them to paste a copied Terminal command, installing macOS malware including Atomic Stealer variants. The ad was later removed after Jamf alerted X and the account owner.
The case shows how paid placement, verification signals, and lookalike domains can be combined to bypass trust checks while keeping execution user-driven. For defenders, the key indicators are sponsored social lures, clipboard-based Terminal prompts, and brand impersonation tied to fake utility downloads.
🛰️ Open sources - closed narratives
@sitreports
🔍 Parrot 7.3 shifts focus to performance and launcher infrastructure
Parrot 7.3 has been released with an opt-in set of optimized builds for x86-64-v3 and ARMv8.2-A, targeting gains in compression, encryption, hashing, media encoding, and other compute-heavy workloads. The release also replaces older shell-based menu handling with two Go binaries, parrot-exec and launcher-updater, while adding official amd64 Vagrant boxes for Home and Security editions.
The update is notable less for new tools than for system-layer changes: higher CPU baselines for selected packages, reduced image bloat, and a faster on-demand launcher path for uninstalled applications. Compatibility-sensitive services remain on conservative baselines, preserving standard apt/dpkg behavior and update flow.
🛰️ Open sources - closed narratives
@sitreports
Parrot 7.3 has been released with an opt-in set of optimized builds for x86-64-v3 and ARMv8.2-A, targeting gains in compression, encryption, hashing, media encoding, and other compute-heavy workloads. The release also replaces older shell-based menu handling with two Go binaries, parrot-exec and launcher-updater, while adding official amd64 Vagrant boxes for Home and Security editions.
The update is notable less for new tools than for system-layer changes: higher CPU baselines for selected packages, reduced image bloat, and a faster on-demand launcher path for uninstalled applications. Compatibility-sensitive services remain on conservative baselines, preserving standard apt/dpkg behavior and update flow.
🛰️ Open sources - closed narratives
@sitreports
🔍 MFA-optional banks keep a major access path exposed
A reported US fraud case saw attackers reuse compromised or recycled credentials to enter bank, retirement, and Gmail accounts, then hide alerts with mailbox filters and move roughly $30,000 before recovery. The broader issue outlined in MFA-optional banks is that several major institutions still leave multi-factor authentication disabled by default or optional.
Operationally, optional MFA creates a predictable weak point: attackers only need one password-based entry path, especially on web logins where app biometrics do not apply. Convenience-driven authentication policy directly expands account takeover risk and delays fraud visibility.
🛰️ Open sources - closed narratives
@sitreports
A reported US fraud case saw attackers reuse compromised or recycled credentials to enter bank, retirement, and Gmail accounts, then hide alerts with mailbox filters and move roughly $30,000 before recovery. The broader issue outlined in MFA-optional banks is that several major institutions still leave multi-factor authentication disabled by default or optional.
Operationally, optional MFA creates a predictable weak point: attackers only need one password-based entry path, especially on web logins where app biometrics do not apply. Convenience-driven authentication policy directly expands account takeover risk and delays fraud visibility.
🛰️ Open sources - closed narratives
@sitreports
🔍 EY fires two staff over alleged access to Australian PM bank data
EY Australia has dismissed two staff accused of accessing bank account details linked to Prime Minister Anthony Albanese while working on a Commonwealth Bank contract. The bank has not explained how contractors were able to reach such sensitive records. The case was outlined in EY Australia coverage published 6 July.
The incident adds another data-handling failure to Australia’s major consultancy sector. Operationally, it highlights the exposure created when external contractors are embedded inside financial workflows with access paths to high-value personal and political data.
🛰️ Open sources - closed narratives
@sitreports
EY Australia has dismissed two staff accused of accessing bank account details linked to Prime Minister Anthony Albanese while working on a Commonwealth Bank contract. The bank has not explained how contractors were able to reach such sensitive records. The case was outlined in EY Australia coverage published 6 July.
The incident adds another data-handling failure to Australia’s major consultancy sector. Operationally, it highlights the exposure created when external contractors are embedded inside financial workflows with access paths to high-value personal and political data.
🛰️ Open sources - closed narratives
@sitreports
🤖 Flipper Zero shifts firmware model to community-led maintenance
Flipper Devices says Flipper Zero firmware will continue to be maintained, but full-time feature development has ended. The company says the codebase reached maturity after the 1.0 release in 2024 and stable version 1.4.3 in December 2025. Requests will be reviewed weekly through GitHub Discussions, with stricter pull-request checks and mandatory integration and regression testing.
The move formalizes a narrower core-team role: oversight, review, and quality control rather than active expansion. It also centralizes user demand into a vote-based workflow, reducing direct support load while keeping the firmware operational as the company shifts resources to new hardware.
🛰️ Open sources - closed narratives
@sitreports
Flipper Devices says Flipper Zero firmware will continue to be maintained, but full-time feature development has ended. The company says the codebase reached maturity after the 1.0 release in 2024 and stable version 1.4.3 in December 2025. Requests will be reviewed weekly through GitHub Discussions, with stricter pull-request checks and mandatory integration and regression testing.
The move formalizes a narrower core-team role: oversight, review, and quality control rather than active expansion. It also centralizes user demand into a vote-based workflow, reducing direct support load while keeping the firmware operational as the company shifts resources to new hardware.
🛰️ Open sources - closed narratives
@sitreports
📡 Iran-linked cluster deploys Cavern C2 against Israeli targets
Researchers have identified a new command-and-control framework, Cavern C2, used by an Iran-linked threat actor in operations targeting Israeli organizations. The campaign adds a fresh malware management layer to an already active regional cyber confrontation.
A dedicated C2 framework matters because it improves operator control, tasking, and persistence across multiple intrusions. Its emergence indicates continued tool development rather than one-off reuse, giving defenders a new infrastructure and tradecraft set to map, detect, and disrupt.
🛰️ Open sources - closed narratives
@sitreports
Researchers have identified a new command-and-control framework, Cavern C2, used by an Iran-linked threat actor in operations targeting Israeli organizations. The campaign adds a fresh malware management layer to an already active regional cyber confrontation.
A dedicated C2 framework matters because it improves operator control, tasking, and persistence across multiple intrusions. Its emergence indicates continued tool development rather than one-off reuse, giving defenders a new infrastructure and tradecraft set to map, detect, and disrupt.
🛰️ Open sources - closed narratives
@sitreports
🔍 16-year-old KVM flaw enables guest-to-host escape on x86
A newly disclosed Linux KVM vulnerability affects Intel and AMD x86 systems and allows a guest virtual machine to escape isolation and execute on the host. The issue reportedly persisted in KVM for 16 years, turning a core virtualization boundary into a potential host-level compromise path. Technical details are outlined in the KVM flaw coverage.
The significance is direct: multi-tenant and virtualized environments depend on KVM isolation as a primary security control. A guest-to-host escape collapses that assumption and can expose the hypervisor layer, with immediate relevance for cloud infrastructure, enterprise virtualization, and lab environments.
🛰️ Open sources - closed narratives
@sitreports
A newly disclosed Linux KVM vulnerability affects Intel and AMD x86 systems and allows a guest virtual machine to escape isolation and execute on the host. The issue reportedly persisted in KVM for 16 years, turning a core virtualization boundary into a potential host-level compromise path. Technical details are outlined in the KVM flaw coverage.
The significance is direct: multi-tenant and virtualized environments depend on KVM isolation as a primary security control. A guest-to-host escape collapses that assumption and can expose the hypervisor layer, with immediate relevance for cloud infrastructure, enterprise virtualization, and lab environments.
🛰️ Open sources - closed narratives
@sitreports
📡 TrojPix turns HDMI cables into covert transmitters
Researchers presented TrojPix, an EM exfiltration method that encodes data into imperceptible pixel changes on digital displays. Malware with only user-level access can modulate TMDS emissions from standard video cables, reaching 8.1 Mbps at up to 208 meters. Tests across nine monitor brands and 15 cables reportedly achieved near-perfect recovery, including through a 30 cm concrete wall.
The operational takeaway is that air gaps remain vulnerable through routine display infrastructure, without hardware modification or visible screen artifacts. The technique expands the threat model for military, industrial, government, and financial networks where HDMI-class links are treated as passive components.
🛰️ Open sources - closed narratives
@sitreports
Researchers presented TrojPix, an EM exfiltration method that encodes data into imperceptible pixel changes on digital displays. Malware with only user-level access can modulate TMDS emissions from standard video cables, reaching 8.1 Mbps at up to 208 meters. Tests across nine monitor brands and 15 cables reportedly achieved near-perfect recovery, including through a 30 cm concrete wall.
The operational takeaway is that air gaps remain vulnerable through routine display infrastructure, without hardware modification or visible screen artifacts. The technique expands the threat model for military, industrial, government, and financial networks where HDMI-class links are treated as passive components.
🛰️ Open sources - closed narratives
@sitreports
📡 Pentagon issues $80.5M counter-UAS task order for Air Force base defense
AeroVironment received an $80.5 million task order to deliver AI-enabled counter-drone systems for Air Force Global Strike Command under the Army-led Domestic Shield initiative. The package includes Titan-MS, Titan 4, EO/IR payloads, and radar for layered defense against small UAS. It is the first order under a broader $500 million contract vehicle announced last week.
The award signals a shift from pilot efforts to fielded, interoperable base-defense architecture focused on detect-and-defeat coverage at fixed installations. Its relevance is amplified by recent unauthorized drone incursions reported at Barksdale AFB, a key AFGSC site.
🛰️ Open sources - closed narratives
@sitreports
AeroVironment received an $80.5 million task order to deliver AI-enabled counter-drone systems for Air Force Global Strike Command under the Army-led Domestic Shield initiative. The package includes Titan-MS, Titan 4, EO/IR payloads, and radar for layered defense against small UAS. It is the first order under a broader $500 million contract vehicle announced last week.
The award signals a shift from pilot efforts to fielded, interoperable base-defense architecture focused on detect-and-defeat coverage at fixed installations. Its relevance is amplified by recent unauthorized drone incursions reported at Barksdale AFB, a key AFGSC site.
🛰️ Open sources - closed narratives
@sitreports
📡 Marine Corps centralizes drone and counter-drone training
The U.S. Marine Corps has established the Marine Corps Robotics Integration Group at Twentynine Palms to oversee Group 1 and 2 UAS and counter-UAS training. The unit will manage pilot courses, curriculum standardization, instructor qualification, regional coordination, lessons learned, and service-wide training updates.
The move creates a single institutional channel for small-drone and counter-drone training across the force. It shifts these capabilities from dispersed experimentation toward standardized doctrine, formal instruction, and centralized oversight, indicating sustained Marine Corps investment in tactical UAS integration.
🛰️ Open sources - closed narratives
@sitreports
The U.S. Marine Corps has established the Marine Corps Robotics Integration Group at Twentynine Palms to oversee Group 1 and 2 UAS and counter-UAS training. The unit will manage pilot courses, curriculum standardization, instructor qualification, regional coordination, lessons learned, and service-wide training updates.
The move creates a single institutional channel for small-drone and counter-drone training across the force. It shifts these capabilities from dispersed experimentation toward standardized doctrine, formal instruction, and centralized oversight, indicating sustained Marine Corps investment in tactical UAS integration.
🛰️ Open sources - closed narratives
@sitreports
🔍 Veeam backup flaw allows domain user to reach SYSTEM
Veeam disclosed CVE-2026-44963, a critical RCE affecting domain-joined Backup & Replication 12.x servers. Any authenticated domain user can execute code on the Backup Server via the .NET Remoting service on TCP/8000, leading to SYSTEM-level compromise. Builds through 12.3.2.4465 are affected; the issue is patched in 12.3.2.4854. Version 13.x is not affected.
The exposure is operationally significant because the Backup Server centralizes privileged access to backup catalogs, repositories, hypervisors, and cloud targets. A low-privilege domain foothold can therefore be converted into control over recovery infrastructure, including deletion or tampering of restore points.
🛰️ Open sources - closed narratives
@sitreports
Veeam disclosed CVE-2026-44963, a critical RCE affecting domain-joined Backup & Replication 12.x servers. Any authenticated domain user can execute code on the Backup Server via the .NET Remoting service on TCP/8000, leading to SYSTEM-level compromise. Builds through 12.3.2.4465 are affected; the issue is patched in 12.3.2.4854. Version 13.x is not affected.
The exposure is operationally significant because the Backup Server centralizes privileged access to backup catalogs, repositories, hypervisors, and cloud targets. A low-privilege domain foothold can therefore be converted into control over recovery infrastructure, including deletion or tampering of restore points.
🛰️ Open sources - closed narratives
@sitreports