Search This Blog

Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Showing posts with label vulnerabilities and Exploits. Show all posts

Six U-Boot Vulnerabilities Could Enable Pre-Boot Code Execution and Persistent Firmware Attacks

 



Security researchers have identified six vulnerabilities in the widely deployed U-Boot bootloader that could allow attackers to execute malicious code during the earliest stages of a device's startup process. If successfully exploited, the flaws could enable firmware-level attacks capable of bypassing security protections before the operating system loads and establishing malware designed to remain on affected systems.

As one of the most widely used open-source bootloaders, U-Boot plays a fundamental role in the startup sequence of embedded Linux devices by initializing hardware and loading the operating system. It is integrated into a broad range of technologies, including enterprise server Baseboard Management Controllers (BMCs), networking equipment, industrial control systems, Internet of Things (IoT) devices, and numerous other embedded appliances.

Because the bootloader executes before the operating system and endpoint security tools become active, vulnerabilities at this stage can have far-reaching consequences. An attacker who gains control during the boot process may be able to interfere with the system's trusted startup sequence before conventional security controls have an opportunity to detect or prevent malicious activity.

One of U-Boot's primary security mechanisms is Verified Boot, which uses cryptographic signatures to verify the authenticity of firmware and operating system images before they are executed. During startup, only images signed with a trusted cryptographic key are intended to be loaded, helping prevent unauthorized or modified firmware from running on the device.

In a technical report published this week, firmware security company Binarly disclosed six vulnerabilities affecting U-Boot's Flattened Image Tree (FIT) signature verification code. The FIT framework is responsible for validating firmware images during the boot process, making it a critical component of the platform's chain of trust.

According to Binarly, researchers examined the verification logic because of its importance in maintaining firmware integrity during startup. Their analysis uncovered six distinct vulnerabilities ranging from denial-of-service conditions that can interrupt the boot process to flaws capable of enabling arbitrary code execution while processing untrusted firmware images.

The researchers said two of the vulnerabilities could potentially allow arbitrary code execution during firmware verification, while the remaining four can be exploited to trigger crashes during the boot process. Since these weaknesses affect the validation of firmware before the operating system starts, a successful exploit could allow malicious instructions to execute before higher-level security mechanisms become operational.

The disclosed vulnerabilities include a flaw identified as BRLY-2026-037 that can cause U-Boot to crash when processing a specially crafted firmware image and, under certain conditions, may also permit arbitrary code execution. BRLY-2026-038 is a memory corruption vulnerability that could enable attackers to execute malicious code during firmware signature verification. BRLY-2026-039 involves an out-of-bounds read that may force U-Boot to access memory beyond the firmware image, resulting in a system crash. BRLY-2026-040 is a null pointer dereference vulnerability that allows crafted firmware images to terminate the bootloader unexpectedly. BRLY-2026-041 stems from insufficient validation of externally stored firmware data and can also be used to crash vulnerable systems. The sixth flaw, BRLY-2026-042, involves unbounded recursion that can exhaust available stack memory and prevent the bootloader from completing the startup process.

Binarly noted that much of the affected code has been present since U-Boot version 2013.07, meaning the vulnerabilities could impact more than 50 stable releases of the project. Because many hardware manufacturers maintain customized downstream versions of U-Boot within their own firmware, the potential exposure extends beyond the upstream project to a large number of commercial products deployed across multiple industries.

If the arbitrary code execution vulnerabilities are successfully exploited, attackers could gain execution during one of the earliest phases of system initialization. Operating at this level may allow threat actors to alter the boot sequence, disable firmware security mechanisms, deploy persistent firmware malware, or perform other privileged actions before the operating system begins loading.

Firmware-based attacks can also be considerably more difficult to identify than malware operating within the operating system. Since malicious activity occurs before the operating system initializes, traditional endpoint security software and many monitoring tools may have limited visibility into the compromise, allowing malicious modifications to remain undetected for extended periods.

Binarly also noted that exploitation does not necessarily require physical access to a device. Systems equipped with Baseboard Management Controllers that support remote firmware updates could become vulnerable if an attacker first compromises the management interface. In such cases, a specially crafted firmware image could be uploaded and processed during the update process, potentially triggering the identified vulnerabilities.

The researchers reported all six vulnerabilities to the U-Boot maintainers and submitted patches addressing each issue. Those fixes have since been accepted into the project's upstream codebase. However, because U-Boot is incorporated into firmware by individual hardware manufacturers, vendors must integrate the patches into their own firmware releases before updates become available to customers.

Organizations operating embedded systems should monitor firmware advisories issued by their hardware vendors and apply security updates as they become available. Restricting access to firmware management interfaces, securing remote administration services such as BMCs, and verifying firmware authenticity before deployment can further reduce exposure while patches are being distributed.

Devices that have reached end-of-life or no longer receive firmware updates may remain permanently vulnerable, underscoring the long-term security challenges posed by legacy embedded systems that continue operating long after vendor support has ended.

BeyondTrust Patches Four Vulnerabilities in Remote Support and PRA

 




BeyondTrust has released security updates to remediate four vulnerabilities affecting its Remote Support (RS) and Privileged Remote Access (PRA) solutions, including two Critical authentication bypass flaws that could allow attackers to gain unauthorized access to vulnerable appliances under specific deployment configurations. The products are commonly used by organizations to deliver remote technical support and manage privileged access to enterprise systems, making them attractive targets because they often provide administrative access to critical IT environments.

The most severe issues originate within the products' authentication mechanisms, which verify user identities before granting access. Because the vulnerabilities can be triggered before the authentication process is completed, successful exploitation may allow attackers to bypass an important security control without first supplying valid credentials.

One of the Critical vulnerabilities, tracked as CVE-2026-40138, carries a CVSS score of 9.2 and affects both BeyondTrust Remote Support and Privileged Remote Access. According to the advisory, the flaw stems from improper validation of authentication data within the authentication subsystem. Under specific authentication configurations, a network-positioned attacker could bypass access controls and obtain unauthorized access to the appliance, including accounts with elevated privileges.

BeyondTrust also addressed CVE-2026-40139, another Critical vulnerability assigned a CVSS score of 9.2 that impacts Remote Support. The issue results from improper processing of authentication requests and could enable an unauthenticated remote attacker to circumvent authentication controls and gain unauthorized access to affected appliances, including privileged accounts. Similar to CVE-2026-40138, exploitation depends on a particular authentication configuration being enabled, meaning the exposure varies according to how affected environments are deployed.

In addition to the authentication bypass flaws, the company disclosed CVE-2026-40140, a High-severity vulnerability with a CVSS score of 8.7 affecting the network communication subsystem. The issue arises from insufficient validation of client-supplied input and could allow an unauthenticated remote attacker to trigger a denial-of-service (DoS) condition, disrupting the availability of vulnerable appliances rather than providing direct access to them.

The fourth vulnerability, CVE-2026-40141, received a CVSS score of 8.5 and affects web application components within both Remote Support and Privileged Remote Access. Caused by inadequate validation of user-supplied input, the flaw could enable an authenticated user with limited privileges to access resources or information beyond their intended authorization. BeyondTrust noted that exploitation of this vulnerability is limited to accounts that already possess specific permissions.

The company said the vulnerabilities were identified during ongoing internal security assessments with assistance from publicly available artificial intelligence models, including Anthropic Claude Opus 4.8, alongside BeyondTrust's proprietary security research tooling. The use of AI-supported analysis reflects a growing trend of incorporating large language models into vulnerability research to assist security teams in identifying potential weaknesses alongside conventional testing techniques.

According to BeyondTrust, the most severe vulnerabilities could allow authentication bypass and unauthorized access when affected systems are configured in specific ways. The remaining flaws could result in service disruption, unintended access to data, or expanded privileges for authenticated users under defined conditions, potentially affecting the confidentiality, availability, and integrity of vulnerable systems.

The vulnerabilities have been resolved in Remote Support version 25.3.3 and later and Privileged Remote Access version 25.3.3 and later. Organizations running version 25.3.2 or earlier of either product are advised to upgrade to the latest available release to mitigate the disclosed risks.

BeyondTrust stated that it has not observed evidence of the newly disclosed vulnerabilities being exploited in the wild. Nevertheless, the company noted that its Remote Support and Privileged Remote Access products have previously been targeted by threat actors. Earlier vulnerabilities, including CVE-2024-12356 and CVE-2026-1731, were exploited to deploy web shells and backdoors on compromised appliances, demonstrating the continued interest of attackers in enterprise remote access infrastructure. Given that history and the privileged role these products play within enterprise environments, organizations are encouraged to apply the available security updates promptly to reduce their exposure to potential attacks.

Ivanti Patches New EPMM Vulnerability Linked to Active Zero-Day Exploitation

 



Software provider Ivanti has released security updates for a newly identified vulnerability in its Endpoint Manager Mobile (EPMM) platform after confirming that the flaw has already been used in limited zero-day attacks.

The vulnerability, tracked as CVE-2026-6973, has been classified as high severity. According to Ivanti, the issue is caused by improper input validation, which refers to a weakness in how an application processes and checks incoming data before handling a request. If exploited successfully, the flaw could allow a remote attacker with administrator-level access to run arbitrary code on vulnerable systems.

Ivanti stated that the vulnerability affects EPMM version 12.8.0.0 and earlier releases. To reduce exposure, the company has issued patched versions including EPMM 12.6.1.1, 12.7.0.1, and 12.8.0.1. The company is also advising customers to review accounts with administrative privileges and rotate credentials where necessary, particularly in environments where earlier compromise activity may have occurred.

In its advisory, Ivanti said the exploitation activity observed so far appears to be limited in scope and requires valid administrator authentication in order to succeed. The company added that it has not identified active exploitation involving the additional vulnerabilities disclosed alongside CVE-2026-6973.

Ivanti also clarified that the issue impacts only the on-premises version of Endpoint Manager Mobile. The company said the flaw does not affect Ivanti Neurons for MDM, which is its cloud-based endpoint management platform. Other products, including Ivanti EPM and Ivanti Sentry, were also listed as unaffected.

Data published by internet monitoring organization Shadowserver Foundation currently shows more than 850 internet-accessible IP addresses associated with Ivanti EPMM deployments. Most of the exposed systems appear to be located in Europe, followed by North America. However, there is still no public visibility into how many of those servers have already installed the latest patches.

Alongside the actively exploited flaw, Ivanti disclosed fixes for four additional high-severity vulnerabilities identified as CVE-2026-5786, CVE-2026-5787, CVE-2026-5788, and CVE-2026-7821. According to the company, these flaws could potentially be used to obtain administrator access, impersonate registered Sentry hosts to receive valid certificate authority-signed client certificates, invoke unauthorized methods, or gain access to restricted information stored within affected environments.

The company stated that it currently has no evidence showing these four vulnerabilities have been exploited in real-world attacks. Ivanti also noted that CVE-2026-7821 affects only organizations using Apple Device Enrollment configurations.

The latest disclosure follows earlier security incidents involving Ivanti EPMM earlier this year. In January, the company disclosed two separate code-injection vulnerabilities, tracked as CVE-2026-1281 and CVE-2026-1340, which were also exploited as zero-days against what Ivanti described at the time as a very limited number of customers.

Ivanti now says customers who followed its earlier recommendation to rotate credentials after the January incidents are likely to face a significantly lower risk of exploitation from CVE-2026-6973. The guidance reflects a growing concern within the cybersecurity industry that attackers often attempt to reuse stolen administrative credentials across multiple intrusion campaigns.

The issue also drew attention from the U.S. Cybersecurity and Infrastructure Security Agency earlier this year. In April, the agency instructed federal civilian agencies to secure vulnerable systems against attacks involving CVE-2026-1340 within four days after adding the flaw to its Known Exploited Vulnerabilities catalog.

Ivanti products have repeatedly appeared in incident response investigations over the last several years, particularly because endpoint and device management platforms typically operate with elevated privileges across enterprise networks. Security agencies and researchers have warned that these systems remain attractive targets for threat actors seeking broad administrative control over organizational infrastructure.

According to data previously published by CISA, 33 Ivanti vulnerabilities have been publicly identified as exploited in the wild, including 12 that were also linked to ransomware-related activity.

Ivanti says it currently serves more than 40,000 customers worldwide through a partner network consisting of over 7,000 organizations.