CISO

Security Advisory

Critical Lua Sandbox Escape Flaw in Redis Allows Remote Code Execution (RCE)

Summary: Security Advisory: A critical vulnerability has been found in the Lua scripting engine of Redis, enabled by default in all versions, allows authenticated attackers to break out of the Lua sandbox and perform remote code execution (RCE) to gain full control of the affected system.

OEMRedis
SeverityCritical
CVSS Score10.0
CVEsCVE-2025-49844
POC AvailableYes
Actively ExploitedNo
Exploited in WildNo
Advisory Version1.0

Overview

Since Redis is used in most cloud environments the impact is highly critical. Redis team has released the patches and urged for immediate updates recommended to secure systems.

Vulnerability NameCVE IDProduct AffectedSeverityFixed Version
Lua Use-After-Free RCE Vulnerability  CVE-2025-49844All Redis Software & OSS/CE/Stack versions with Lua scripting  CriticalRedis Software: 7.22.2-12+, 7.8.6-207+, 7.4.6-272+, 7.2.4-138+, 6.4.2-131+ Redis OSS/CE: 8.2.2+, 8.0.4+, 7.4.6+, 7.2.11+ Redis Stack: 7.4.0-v7+, 7.2.0-v19+

Technical Summary

The vulnerability comes from a use-after-free (UAF) bug in Redis’s Lua scripting system, caused by improper checks during memory cleanup. Authenticated attackers can send malicious Lua scripts via EVAL or EVALSHA commands to manipulate memory, bypass the sandbox, and run arbitrary code. Even internal servers are at risk if attackers gain network access, making this flaw highly critical for both exposed and internal environments.

CVE IDSystem AffectedVulnerability DetailsImpact
CVE-2025-49844All Redis Software & OSS/CE/Stack below the fixed versionA user after free in the Lua garbage collector allows memory corruption via crafted scripts, enabling sandbox escape and RCERemote Code Execution

Recommendations

Upgrade to the below  fixed versions immediately.

  • Redis Software: 7.22.2-12+, 7.8.6-207+, 7.4.6-272+, 7.2.4-138+, 6.4.2-131+
  • Redis OSS/CE: 8.2.2+, 8.0.4+, 7.4.6+, 7.2.11+
  • Redis Stack: 7.4.0-v7+, 7.2.0-v19+

Here are some best practices

  • Enable Strong Authentication: Configure strong passwords on all the instances, ensure protected-mode is enabled (in CE and OSS) to prevent accidental exposure.
  • Network Controls: Restrict access to authorized IPs using firewalls or VPCs, limit access to trusted sources and prevent unauthorized connectivity.
  • Limit permissions: To enhance security, user needs to give minimum necessary permissions.
  • Monitoring: Check the logs to see if there are any suspicious activities.
  • Incident Response: If compromised, isolate systems, rotate credentials, and scan for malware.

Conclusion:
This is a critical vulnerability with a CVSS score of 10.0, affecting all Redis versions with Lua scripting. The widespread Redis usage, default insecure configurations makes this a critical threat. Immediate patching and hardening are essential to prevent full system compromise, data breaches, and further attacks.

References:

Blogs

DoW Announced Implementation of CSRMC to Deliver Real Time Cyber Defense, Address Legacy Shortcomming’s

Managing cyber risk across the cyber security set up of an enterprise is harder than ever and keeping architectures and systems secure also compliant can be challenging and over whelming.

DoW (Deprtament of war) recently announced implementing of a groundbreaking Cybersecurity Risk Management Construct (CSRMC).

This is a transformative framework to deliver real-time cyber defense at operational speed and its five-phase construct that ensures a hardened, verifiable, continuously monitored and actively defended environment to ensure that U.S. warfighters maintain technological superiority against rapidly evolving cyber threats.

In comparison the previous Risk management framework dependent on static checklists and manual processes . The framework failed to account for operational needs and cyber survivability requirements. 

How (CSRMC) is going to address legacy infrastructure shortcoming?

CSRMC addresses these gaps by shifting from “snapshot in time” assessments to dynamic, automated, and continuous risk management, enabling cyber defense at the speed of relevance required for modern warfare.

The construct is composed of a five-phase lifecycle and ten foundational tenets.

The Five-Phase Lifecycle

The new construct organizes cybersecurity into five phases aligned to system development and operations:

  1. Design Phase – Security is embedded at the outset, ensuring resilience is built into system architecture.
  2. Build Phase – Secure designs are implemented as systems achieve Initial Operating Capability (IOC).
  3. Test Phase – Comprehensive validation and stress testing are performed prior to Full Operating Capability (FOC).
  4. Onboard Phase – Automated continuous monitoring is activated at deployment to sustain system visibility.
  5. Operations Phase – Real-time dashboards and alerting mechanisms provide immediate threat detection and rapid response.

Ten Foundational Tenets

The CSRMC has 10 core principal

  • Automation – driving efficiency and scale
  • Critical Controls – identifying and tracking the controls that matter most to cybersecurity
  • Continuous Monitoring and ATO – enabling real-time situational awareness to achieve constant ATO posture
  • DevSecOps – supporting secure, agile development and deployment
  • Cyber Survivability – enabling operations in contested environments
  • Training – upskilling personnel to meet evolving challenges
  • Enterprise Services & Inheritance – reducing duplication and compliance burdens
  • Operationalization – ensuring stakeholders near real-time visibility of cybersecurity risk posture
  • Reciprocity – reuse assessments across systems
  • Cybersecurity Assessments – integrating threat-informed testing to validate security

“This construct represents a cultural fundamental shift in how the Department approaches cybersecurity,” said Kattie Arrington, performing the duties of the DoW CIO. “With automation, continuous monitoring, and resilience at its core, the CSRMC empowers the DoW to defend against today’s adversaries while preparing for tomorrow’s challenges.”

With the above tenants DoW is ensuring cyber survivability and mission assurance in every domain,air, land, sea, space, and cyberspace.

Addressing Cyber security risk management

Cybersecurity risk management isn’t simply the job of the security team; everyone in the organization has a role to play. Often siloed, employees and business unit leaders view risk management from their business function.

Intru360 gives security analysts and SOC managers a clear view across the organization, helping them fully understand the extent and context of an attack. It also simplifies workflows by automatically handling alerts, allowing for faster detection of both known and unknown threats.

Identify latest threats without having to purchase, implement, and oversee several solutions or find, hire, and manage a team security analyst. Unify latest threat intelligence and security technologies to prioritize the threats that pose the greatest risk to your company.

GaarudNode is an all-in-one  solution designed to empower development teams with the tools they need to secure their applications throughout the development lifecycle. By combining the power of SAST, DAST, SCA, API security, and CSPM, GaarudNode provides a comprehensive security framework that ensures your applications are built, tested, and deployed with confidence.

Mirage Cloak offers various deception methods to detect and stop threats before they cause damage. These methods include adding decoys to the network, deploying breadcrumbs on current enterprise assets, using baits as tripwires on endpoints, and setting up lures with intentionally misconfigured or vulnerable services or applications. The flexible framework also lets customers add new deception methods as needed.

BISO Analytics stands out as the pioneering security analytics platform designed to assist enterprises in effectively handling their first-party, third-party, and emerging risks, all within a single platform. This comprehensive solution facilitates a quicker and safer progression for your business.

By adopting a groundbreaking approach, BISO Analytics integrates open, data-centric cyber risk management practices, offering organizations a consolidated view of their cyber risk landscape across the entire attack surface.

BISO Analytics empowers CXO, mid-management, and operational teams with real-time, reliable, and defensible data that not only complies with regulatory standards but also aligns with the expectations of the board regarding safeguarding shareholder value and fortifying the business.

Why it is important to implement cybersecurity risk management at organisational level

Having an effective cybersecurity risk management program can only be implemented in an organization through a structured process. This requires careful planning, resource allocation and commitment to improving security framework.

Registering documents that assess risk related activities include high asset inventories like all systems and data. When risk are registered it contain records of determined risk, data theft or results of assessment and planned treatments.

Organizations that possess all documentation involving controls and their implementation level. In this scenario organizations actually understands what exactly is risk assessment and identifying what can go wrong in an organization’s system either anything that is via threats, vulnerabilities and their possible impact.

As the saying goes we can’t protect what you don’t understand and one can’t manage what they don’t assess.

Visit our website for more informed details on our products.

(Source: www.miragenews.com/war-dept-unveils-new-cybersecurity-risk-1540279/)

Blogs

Discord Security Incident Reveal Support Ticket Stolen in Third-Party Breach 

Summary 

In today’s interconnected digital world, trust often reaches beyond the main platforms to include the network of partners that support them. Recently, Discord disclosed an incident tied not to its own systems, but to a third-party customer service provider whose systems were compromised, exposing limited user information.

The company emphasized that its core infrastructure remained secure, with the issue confined to the external vendor’s network. The cyber attack appears to be financially motivated, as hackers demanded a ransom from Discord in exchange not to reveal any information that they ceased.

Breach Breakdown 

Discord confirmed that an unauthorized party gained access to the vendor’s systems by exploiting that firm’s ticketing platform. Through that entry point, attackers were able to view limited user information exchanged during support requests like names, Discord handles, emails and some contact details.

For a small number of affected users, the exposure extended to scanned government-issued IDs, such as passports or driver’s licenses, typically used to verify age or ownership. 

Discord as well clarified that its core infrastructure like chat servers, authentication databases, and private messaging systems were not breached. This wasn’t a platform intrusion, but rather a supply chain compromise through one of the company’s external service tools. 

Upon discovering the breach, Discord revoked the vendor’s access immediately, launched an internal investigation. Law enforcement agencies have also joined the effort to identify the perpetrators and prevent further misuse of stolen data. 

Discord already notified data protection authorities, contacted affected users directly via noreply@discord.com, and reviewed all vendor relationships to ensure compliance with data protection standards. The company also pledged to strengthen third-party systems and increase security oversight for partners with data access.

Additionally, Discord advised users to watch for phishing attempts and reiterated that it will never contact them by phone regarding the incident. 

Recommendations: 

Here are some recommendations below 

  • Always verify the sender before clicking links in security emails. 
  • Enable multifactor authentication to protect your account even if credentials leak. 
  • Stay alert for phishing emails, especially those that sound urgent or official. 
  • Keep your data footprint minimal by sharing only what’s necessary. 
  • Regularly assess vendor security and treat third-party reviews as a key defense measure. 

Conclusion 
This incident underscores that even well-secured platforms like Discord remain vulnerable through their third-party partners. It highlights the growing importance of robust vendor risk management, transparent communication, and continuous security auditing.

For users, it’s a reminder to stay cautious, enable strong authentication measures, and practice vigilance against phishing or social engineering attempts following any major data disclosure. 

Discord was created as a communication platform for gamers, who represent more than 90% of the userbase, but expanded to various other communities, allowing text messages, voice chats and video calls.

References: 

Blogs

Red Hat Hit by Data Breach exposing major sensitive data, including the NSA

Red Hat, has been allegedly been hit by a breach and this has been posted by Crimson Collective hackers group on Telegram. The cyber criminals claim they’ve snatched private GitHub repositories, which include sensitive data about approximately 800 customers’ networks.

Key points from the RedHat Breach

  • Data from 28,000 internal projects at Red Hat has allegedly been stolen.
  • The hacker group Crimson Collective claims to have stolen nearly 570GB of data.
  • Extortion group known as Crimson Collective posted of they gaining access to over 28,000 Red Hat repositories, containing 570.2 GB in total.
  • The data extracted data allegedly includes around 800 Customer Engagement Reports (CERs), exposing sensitive customer information, such as their network configurations.
  • The hackers posted the claims on a Telegram channel created on September 24th, 2025. As proof, the cybercriminals provided the entire file tree, a list of allegedly stolen CERs, and some other screenshots.
  • According to International Cyber Digest, these include the National Security Agency (NSA), the Department of Energy, the National Institute of Standards and Technology (NIST), IBM, Citi, Verizon, Siemens, Bosch, JPMC, HSBC, Telefonica, other major telecoms, banks, and many other organizations.

“Source code and consulting engagement reports (CERs), if leaked, can help attackers analyze internal company infrastructure and software running on that infrastructure. This makes it significantly easier and faster to identify vulnerable attack vectors for potential attackers, “ said Aras Nazarovas, information security researcher at Cybernews.

RedHat confirmed the attack

According to the attackers, they found authentication keys, full database URIs, and other private information in the Red Hat code and CERs, which they allegedly used to gain access to downstream customer infrastructure.

On Telegram, the hacker group published a complete directory listing of stolen GitHub repositories, along with a list of customer reports from the period 2020-2025.

Red Hat has confirmed the security incident relating to its GitLab instance, but declined to comment on the attackers’ specific claims regarding the GitHub repositories and customer reports. The company emphasizes that there is no reason to believe that the security issue affects other Red Hat services or products. Red Hat says it is very confident in the integrity of its software supply chain.

The CER list includes organizations from various sectors, including major international names such as Bank of America, T-Mobile, AT&T, Fidelity, and Walmart.

Extortion Demands by Hackers

The data breach on RedHat was also an attempt to contact Red Hat and get through with extortion demands. The cybercriminals received a response asking them to submit a vulnerability report to the security team.

The ticket created by cyber criminals was reportedly forwarded repeatedly to various individuals, including employees of Red Hat’s legal and security departments.

Blogs

Service Provider for Volvo NA, ‘Miljödata’ hit by Ransomware; Critical Data exposed

Third-party supplier Miljödata, for Volvo North America,hit by ransomware disclosed a data breach that exposed the personal data of its employees . The ransomware attack happened in month of August 2025. and impacted at least 25 companies. The ransomware group DataCarry claimed responsibility for the attack on Miljödata and also published allegedly stolen data on its Tor leak site.

Ransomware attacks are increasingly targeting both enterprise of all sizes across all sectors. The attack affected Scandinavian airline SAS, Boliden and included 200 Swedish municipalities. The affected systems were mostly for HR purposes that handled medical certificates, rehabilitation matters, reporting and managing work-related injuries.

The service provider of Volvo, launched an investigation into the incident with the help of cybersecurity experts, enhanced the security of its hosted environment, and is working to prevent similar security breaches in the future.

According to the data breach notification service Have I Been Pwned (HIBP), the leaked data belongs to 870,000 accounts. Exposed data includes email addresses, names, physical addresses, phone numbers, government IDs, dates of birth, and gender.

DataCarry Ransomware Group

The DataCarry ransomware group claimed responsibility for the attack on Miljödata’s Adato system, and has Miljödata’s files available for download on its dark web-based site.

Need of the hour for Enterprise security who are soft target of ransomware attack.

  • Continuously monitor to detect breached credentials, leaked databases, and threat actor’s activites in near real-time before damage gas taken full control.
  • Assessment on cyber attack module as soon as an attack was initiated and do proper full incident review to determine how attackers infiltrated enterprise network and how data exfiltrated and if there is any existing threat.
  • Authenticate backups of data that have been stored currently and if they have been encrypted or stored offline. It is responsibility of enterprise to keep immutable backup solutions to defend against any ransomware attack that may encompass from encryption and deletion attempts by threat actors.
  • Implement threat intelligence for real time alert against any external threat that gets feeder into system . Enterprise security must Include indicators of compromise (IOCs), into company’s XDR platforms for real-time alerting .
  • Include phishing simulations and enforce multi-factor authentication (MFA) across all access points.

While Volvo did not specify the exact scale of its breach, it is one of many large organizations to be caught up in the data raid. As per reports Volvo Group provided the affected individuals with 18 months of free identity protection and credit monitoring services.

Source: Volvo North America disclosed a data breach following a ransomware attack on IT provider Miljödata

Blogs Cybersecurity News

Telecom Network in New York Area Dismantled after Network Threat Detected

The US Secret Service, the agency in charge of security for the United Nations General Assembly, discovered a threatening network of over 300 servers and 10,000 SIM cards across the New York tri-state area.

The network could have “disabled cell phone towers and potentially shut down the cellular network in New York City,” Matt McCool, the special agent in charge of the Secret Service’s New York field office.

Key Points:

The network could also facilitate denial of service attacks and could send up to 30 million text messages per minute. All of the devices were found within 35 miles of the United Nations headquarters in Midtown Manhattan.

Analysis indicates cellular communications between nation-state threat actors and individuals that are known to federal law enforcement the report said.

The investigation into the devices is ongoing, the Secret Service said, but early forensic analysis indicates it was used for communications between “foreign actors” and people already known to federal law enforcement. No arrests have been announced, and investigators are still searching through the equivalent of 100,000 cell phones worth of data.

“This network had the potential to disable cell phone towers and essentially shut down the cellular network in New York City,” Matt McCool, special agent in charge of the Secret Service field office in New York, said in a video statement.

The telecommunications gear was recovered from so-called SIM farms housed in abandoned apartment buildings in at least five undisclosed sites. The devices discovered could be used to conduct a range of telecommunications attacks including disabling cell phone towers, enabling cybersecurity attacks and allowing encrypted communication between criminal groups and threat actors.

According to the Secret Service, the devices could facilitate a wide range of attacks on telecommunications systems, including disabling cell phone towers, enabling denial of service attacks.

This also allowed encrypted, anonymous communication between potential threat actors and criminal enterprises.

The forensic analysis indicates potential links between the network and overseas threat actors, as well as connections to individuals already known to federal law enforcement agencies.

According to Bloomberg, it is still unclear whether the network was connected to earlier incidents this year in which unknown individuals impersonated White House Chief of Staff Susie Wiles and Secretary of State Marco Rubio.

A full forensic review of the seized devices is ongoing as authorities continue to assess the scope and origins of the network.

Investigations started after threats to US officials

According to agents who spoke to the New York Times, the investigation began after anonymous telephonic threats were made against three US government officials earlier this year. One of the officials who was threatened worked with the Secret Service, while the other two were White House staffers.

State of crime

The agency first detected the New York-area SIM farm after it was linked to swatting incidents on Christmas Day in 2023. Those incidents involved Congresswoman Marjorie Taylor Greene and US Senator Rick Scott.

The cases were tied to two Romanian men, Thomasz Szabo and Nemanja Radovanovic, who were working with an American swatter, Alan Filion, also known as “Torswats.” All three have since been convicted on swatting-related charges.

Ben Coon, head of intelligence at cybersecurity firm Unit 221b, believes there was little foreign state involvement, and the operation is based on financial crimes.

Images released by the Secret Service showed racks of neatly arranged telecom equipment, each component numbered and labeled. Cables were carefully laid out and secured, which could mean the operation was handled by well-resourced professionals.

The operation is linked to swatting incidents, organized crime groups, and nation-state actors, with equipment seized across New York and New Jersey.

Sources: https://www.telegraphindia.com/world/us-secret-service-dismantles-telecom-threat-network-in-new-york-ahead-of-un-general-assembly/cid/2124609


Blogs

𝐊𝐓 𝐓𝐞𝐥𝐞𝐜𝐨𝐦 𝐁𝐫𝐞𝐚𝐜𝐡 𝐑𝐞𝐯𝐞𝐚𝐥𝐬 𝐡𝐨𝐰 Illegal 𝐁𝐚𝐬𝐞 𝐒𝐭𝐚𝐭𝐢𝐨𝐧𝐬 Generated for 𝐇𝐚𝐜𝐤 𝐩𝐚𝐲𝐦𝐞𝐧𝐭𝐬 

Imagine you come to know small payments via your mobile phone is being carried out without your knowledge & come to know that payments are directed to small base stations created by hackers linking your service providers.

 Cyber criminals hacked ultra-small base stations accessed the KT communication network and intercepted traffic during an on-site inspection on the 8th sep.

The Telcom giant got hacked in a clever managed systematic way when the hacker has created a similar base station by stealing femtocells that are not used or under-managed. KT has disconnected the base station in question.

To prevent a recurrence, it will upgrade the management system for micro base stations and strengthen a system that monitors abnormal payment types in real time. It will convert about 2,000 stores nationwide into “Safe and Secure Specialty Stores” and provide affected customers with the “KT Safe and Secure Insurance” (tentative name) free of charge for the next three years to compensate for financial fraud linked to communication devices.

This happened when KT, the south Korean telecom provider discovered two additional illegal ultrasmall base stations, or femtocells, that were used to facilitate a large-scale micropayment scam, bringing the confirmed total to four.

The telecom giant said Thursday that the devices had leaked IMSI, IMEI and phone numbers, and that number of confirmed impacted subscribers had risen from 278 to 362 and that funds embezzled through fraudulent charges to gift cards and transit passes had reached 240 million won, or 173-thousand U.S. dollars. 

Attacks on devices

KT said no additional funds have been stolen since it blocked abnormal transactions on September 5, and that all newly confirmed cases predate that date.

In this attack type personal details such as names and birth dates were not leaked via its network and that SIM authentication keys remain secure, meaning perpetrators of the data breach do not have the ability to clone impacted users’ devices.

Mitigation steps by KT

KT said it is reimbursing victims, offering free SIM card replacements and instructing customers via its website and app, as well as text message, to keep an eye out for fraudulent charges and sign up for the carrier’s SIM protection service.

To prevent a recurrence, it will upgrade the management system for micro base stations and strengthen a system that monitors abnormal payment types in real time.

It will convert about 2,000 stores nationwide into “Safe and Secure Specialty Stores” and provide affected customers with the “KT Safe and Secure Insurance” (tentative name) free of charge for the next three years to compensate for financial fraud linked to communication devices.

Security Advisory

Radware Uncovers Server Side Attack Targeting ChatGPT Known as Shadowleak

Researchers at Radware uncovered a server-side data theft attack targeting ChatGPT, termed as ShadowLeak. The experts discovered the zero-click vulnerability in ChatGPT’s Deep Research agent when connected to Gmail and browsing. 

In this attack type ‘Service-side’ pose greater risk as enterprise defenses cannot detect exfiltration because it runs from the provider’s infrastructure.

ShadowLeak a Server side attack

For any normal user there would be no visible signs of data loss as the AI agent acts as a trusted proxy, sending sensitive data to attacker-controlled endpoints. These server-side requests face fewer URL restrictions, letting attackers export data to virtually any destination.

Shadowleak is an uncovered security flaw affecting ChatGPT’s Deep Research Agent. Which can connect to services like Gmail to help users analyze their emails.

Attackers could hide invisible instructions in a regular looking email. When the user asked ChatGPT to review their mailbox contents selecting deep research.

Vulnerability Details 

ChatGPT’s Deep Research Agent was vulnerable because it could be tricked into following hidden instructions that were inside a seemingly ordinary email. When users ask the agent to analyze their inbox, any attacker can craft the message with invisible commands and cause AI to leak private data without warning.

These hidden instructions used tricks to fool the AI and get around its built-in safety checks. Some of those tricks included: 

  • Pretending to Have Permission: The prompt told the agent that it had “full authorization” to access outside websites, even though it didn’t. 
  • Hiding the Real Purpose: It disguised the hacker’s website as something safe sounding, like a “compliance validation system.” 
  • Telling the Agent to Keep Trying: If the AI couldn’t reach the attacker’s website the first time, the prompt told it to try again helping it sneak past any temporary protections. 
  • Creating Urgency: The prompt warned the agent that if it didn’t follow the instructions, it might not complete the report properly pushing it to obey. 
  • Hiding the Stolen Info: The agent was told to encode the personal data using Base64, which made the data harder to spot and helped hide the theft. 

After reading the fake email, the agent would go look through the user’s real emails (like HR messages) and find personal info such as full names and addresses.

Without alerting the user, the AI would send that information to the attacker’s server, happening silently in the background, with no warning or visible signs. 

This attack is not limited only to Gmail, also applies to any data sources Deep Research accesses, including Google Drive, Dropbox, Outlook, Teams and more. Any connected service that feeds text into the agent can pose a risk to hidden prompts, making sensitive business data vulnerable to exfiltration. 

Source: radware.com 

Attack Flow 

Step Description 
Malicious Email Crafting Attackers create a legitimate email embedded with hidden, invisible prompt instructions to extract sensitive data. Use social engineering and obfuscation. 
Email Delivery and Receipt The victim receives the email in Gmail without needing to open it; hidden commands are present in the email’s HTML body. 
User Invokes Deep Research The victim asks ChatGPT’s Deep Research Agent to analyze their inbox or specific emails, triggering the agent’s activity. 
Parsing Hidden Instructions The agent reads and interprets the hidden malicious prompt embedded within the attacker’s email. 
Extraction of Sensitive Data Following the instructions, the agent locates and extracts personal information like names and addresses from real emails. 
Data Exfiltration to Attacker The agent uses internal tools to send the extracted, often Base64-encoded data to an attacker-controlled external server. 
Victim Remains Unaware The entire process happens silently on OpenAI’s servers with no visible alerts or client-side traces for the user or admins. 

Why It’s Effective 

This “zero-click” attack happened entirely on OpenAI’s servers, where traditional security tools couldn’t detect or stop it, and victims never saw any warning. OpenAI was informed by radware security team in June 2025 and OpenAI fully patched the issue by September. 

The attack runs silently in a trusted cloud environment, invisible to users and traditional security tools.

It tricks the AI into repeatedly sending encoded sensitive data, bypassing safety checks and ensuring successful data theft. This stealthy, zero-click nature means no user interaction is required, making detection extremely difficult and allowing the attacker to exfiltrate data unnoticed over extended periods. 

Recommendations: 

Here are some recommendations below 

  • Email Sanitization: Normalize and strip hidden or suspicious HTML/CSS elements from emails before they are processed by AI agents. This reduces the risk of hidden prompt injections. 
  • Strict Agent Permissions: Limit AI agent access only to the data and tools necessary for its tasks, minimizing exposure to sensitive information. 
  • Behavior Monitoring: Continuously monitor AI agent actions and behavior in real time to detect anomalies or actions deviating from user intent. 
  • Regular Patch Management: Keep AI tools, connectors and integrated systems up to date with the latest security fixes and improvements. 
  • Awareness and Training: Educate users and administrators about the types of attacks AI agents are vulnerable to, fostering vigilance and quick incident response. 

Conclusion 


The ShadowLeak vulnerability underscores the critical risks posed when powerful AI tools operate without sufficient safeguards. By hiding secret commands inside emails, attackers were able to steal personal information without the user knowing.

This case highlights the need for strong safety measures, including limiting AI access to sensitive information, sanitizing inputs to prevent hidden commands, and continuously monitoring agent behavior to detect anomalies.

As more AI tools are used, it’s important to keep strong security controls and oversight to use these technologies safely and protect sensitive data from new threats. 

References: 

Security Advisory

Jenkins Security Patch Fixed HTTP/2 DoS and Permission Issues  

Security advisory: Jenkins addressed critical security flaws in its built-in HTTP server related to the handling of HTTP/2 connections, where attackers could overwhelm servers causing denial of service. This mainly impacts Jenkins instances running with HTTP/2 enabled, which is not the default setting.

Severity High 
CVSS Score 7.7 
CVEs CVE-2025-5115, CVE-2025-59474, CVE-2025-59475, CVE-2025-59476 
POC Available No 
Actively Exploited No 
Exploited in Wild No 
Advisory Version 1.0 

Overview 

Jenkins, a popular open-source automation server used for building and deploying software, recently patched several high & medium security flaws.

The high severity issue is a Denial-of-Service (DoS) vulnerability that could allow attackers to overwhelm the server and make it stop working properly even without needing to log in.

Other issues included the risk of unauthorized users viewing sensitive configuration information and the possibility of attackers inserting fake log entries to confuse system administrators. Jenkins released updates to fix these issues and strongly recommends users upgrade to the latest versions to stay protected. 

                Vulnerability Name CVE ID Product Affected Severity Fixed Version 
HTTP/2 Denial of Service in bundled Jetty  CVE-2025-5115 Jenkins (bundled Jetty)  High Weekly 2.524+, LTS 2.516.3+ 
Missing permission check – agent names CVE-2025-59474 Jenkins core Medium Weekly 2.528+, LTS 2.516.3+ 
Missing permission check – user profile menu CVE-2025-59475 Jenkins core Medium Weekly 2.528+, LTS 2.516.3+ 
Log Message Injection Vulnerability CVE-2025-59476 Jenkins core Medium Weekly 2.528+, LTS 2.516.3+ 

Technical Summary 

Additionally, permission checks in some user interface areas were incomplete, allowing unauthorized users to access sensitive information such as agent names and configuration details.

There was also a vulnerability in log message processing that could let attackers insert misleading entries to confuse administrators. All the issues are fixed in Jenkins latest version. 

CVE ID System Affected  Vulnerability Details Impact 
CVE-2025-5115 Jenkins instances with embedded Jetty server with HTTP/2 enabled It causes the Jetty server to repeatedly reset HTTP/2 streams (RST_STREAM) in response to malicious or malformed frames, leading to resource exhaustion and potential denial of service.  Denial of service 
CVE-2025-59474 Jenkins automation server Permission check flaw allowing unauthorized users to view Jenkins agent/executor names via the side panel executor’s widget Information Disclosure 
CVE-2025-59475 Jenkins automation server Permission check flaw allowing authenticated users without Overall/Read permission to view sensitive configuration details via the Jenkins user profile dropdown menu. Information Disclosure 
CVE-2025-59476 Jenkins automation server An attacker can inject line breaks into Jenkins log messages, leading to forged or misleading log entries. Misleading administrators 

Remediation: 

  • Users should immediately install the latest, patched version of Jenkins on all servers: 
  • Weekly Release: Update to Jenkins v2.528 or later. 
  • Long-Term Support (LTS): Update to Jenkins v2.516.3 or later 

Here are some recommendations below. 

  • If immediate upgrade is not possible, users should disable HTTP/2 to mitigate the Denial-of-Service vulnerability. 
  • Always keep Jenkins core and plugins up to date with the latest security patches. 
  • Regularly audit and monitor access logs and system activity 
     

Conclusion: 
These security flaws could seriously impact Jenkins users, especially those relying on it for continuous integration and deployment. The DoS vulnerability is particularly dangerous because it can be triggered by anyone over the internet, even if they don’t have an account.

Enterprise admins & users should upgrade immediately to the patched versions or disable HTTP/2 to reduce the risk. Keeping Jenkins up to date and following good security practices along with restricting user permissions and monitoring logs is essential to prevent attacks and maintain the stability and safety of software delivery pipelines. 

References: 

Blogs Security Advisory

Shai-Hulud NPM Supply Chain Attack Expands to 470+ Packages 

Summary: A large-scale malicious campaign, nicknamed the Shai-Hulud attack, has impacted the npm ecosystem with over 500 trojanized packages, including those packages maintained by CrowdStrike. The attack originated from a sophisticated phishing campaign that exploited the fundamental trust relationships within the npm ecosystem. 

The JavaScript ecosystem is under a massive threat following a major supply chain attack. Hence, millions of crypto users and developers are now at risk. With more than a billion of these packages downloaded already, thousands of blockchain wallets and applications could be suffer varying exploits.

  • Malicious NPM updates spread malware that steals and replaces crypto addresses.
  • Developers encouraged developer to cease on-chain operation and inspect HD wallets thoroughly.

The attackers injected malicious scripts that

  • Run secret-scanning tools on developer systems, 
  • Steal GitHub, npm and cloud credentials, 
  • Insert persistent GitHub Actions workflows for long-term access, and 
  • Exfiltrate sensitive data to attacker-controlled endpoints. 

This attack is ongoing and all users of npm packages should take immediate steps to secure tokens, audit their environments and verify package integrity. 

Issue Details 

Initial discovery on September 14, 2025, when suspicious versions of @ctrl/tinycolor and ~40 other packages were flagged. By September 16, the attack had spread to include CrowdStrike-namespaced packages and dozens from @ctrl, @nativescript-community, rxnt, @operato, and others. 

Malware behavior 

  • Downloads and runs TruffleHog, a legitimate secret scanner. 
  • Harvests secrets from local machines and CI/CD agents (npm tokens, GitHub PATs, AWS/GCP cloud keys). 
  • Writes malicious workflows into .github/workflows (shai-hulud-workflow.yml). 
  • Continuously exfiltrates findings to a fixed webhook endpoint or pushes them into new GitHub repos under the victim’s account. 

Attack Flow 

Here are some popular packages with affected versions 

Package Version 
@ctrl/ngx-codemirror 7.0.1, 7.0.2 
@ctrl/tinycolor 4.1.1, 4.1.2 
@crowdstrike/foundry-js 0.19.1, 0.19.2 
@crowdstrike/logscale-dashboard 1.205.1, 1.205.2 
@nativescript-community/sqlite 3.5.2 – 3.5.5 
@nativescript-community/text 1.6.9 – 1.6.13 
@nstudio/nativescript-checkbox 2.0.6 – 2.0.9 
@nstudio/angular 20.0.4 – 20.0.6 
eslint-config-crowdstrike 11.0.2, 11.0.3 
remark-preset-lint-crowdstrike 4.0.1, 4.0.2 

Attack Indicators 

Malicious Workflow Filenames 

  • .github/workflows/shai-hulud-workflow.yml 
  • .github/workflows/shai-hulud.yaml 

Exfiltration Endpoint 

  • hxxps://webhook[.]site/bb8ca5f6-4175-45d2-b042-fc9ebb8170b7 

Hashes of Malicious Payloads 

SHA-256 Hash Notes 
46faab8ab153fae6e80e7cca38eab363075bb524edd79e42269217a083628f09 Large batch, Sept 15–16 
b74caeaa75e077c99f7d44f46daaf9796a3be43ecf24f2a1fd381844669da777 CrowdStrike-related packages burst (Sept 16) 
de0e25a3e6c1e1e5998b306b7141b3dc4c0088da9d7bb47c1c00c91e6e4f85d6 First observed compromise (Sept 14) 
81d2a004a1bca6ef87a1caf7d0e0b355ad1764238e40ff6d1b1cb77ad4f595c3 Sept 14 small burst 
83a650ce44b2a9854802a7fb4c202877815274c129af49e6c2d1d5d5d55c501e ~25 packages, Sept 14 
4b2399646573bb737c4969563303d8ee2e9ddbd1b271f1ca9e35ea78062538db Burst of ~17 packages, Sept 14–15 
dc67467a39b70d1cd4c1f7f7a459b35058163592f4a9e8fb4dffcbba98ef210c Multiple reuse across Sept 15–16 

Recommendations: 

Organizations and developers using npm should take immediate actions: 

  1. Uninstall or downgrade 
    Pin dependencies to known-safe versions until patched releases are confirmed. 
  1. Rotate credentials 
    Immediately revoke and reissue: 
  • npm access tokens 
  • GitHub personal access tokens / org tokens 
  • Cloud credentials (AWS, GCP, Azure) 
  1. Audit systems 
  • Inspect developer machines and CI/CD build agents for signs of the malicious bundle.js. 
  • Check .github/workflows for unauthorized files named “shai-hulud-*”. 
  • Review repositories for suspicious commits or new repos labeled “Shai-Hulud Migration”. 
  1. Monitor and log 
  • Search event logs for unusual npm publish activity. 
  • Investigate GitHub Actions runs designed to exfiltrate secrets. 
  1. Harden pipelines 
  • Pin package versions and use integrity checks (e.g.- lockfiles, checksums). 
  • Limit exposure of sensitive tokens in build environments. 
  • Rotate all build-related secrets regularly. 

 
Conclusion 
This incident is significant compromises in the npm ecosystem, impacting hundreds of widely used packages across various namespaces.

The attackers’ tactics such as credential theft, manipulation of GitHub workflows, and widespread package propagation, highlighting the growing sophistication of modern supply chain attacks.

Developers and organizations are strongly advised to take immediate action by removing affected package versions, rotating any exposed secrets, auditing their build environments and strengthening CI/CD security. Continuous monitoring and rapid response are essential to reducing risk and maintaining trust in open-source software. 

The attack’s browser API-level operation revealed critical blind spots in enterprise security monitoring, particularly for organizations handling cryptocurrency transactions.

References: 

Scroll to top