In this roundup, Tony looks at attacks against Polish water treatment facilities, how AI-directed attacks failed in Mexico, and what Google believes is the first AI-generated zero-day exploit
Written by: Takahiro Sugiyama, Peter Revelant, Mathew Potaczek Introduction In late 2025, Mandiant responded to a security incident involving a compromised web server running KnowledgeDeliver. KnowledgeDeliver is a Learning Management System (LMS) developed by Digital Knowledge commonly used in Japan. Mandiant identified a critical vulnerability that allowed unauthenticated Remote Code Execution (RCE). An unknown threat actor leveraged this access to inject malicious code into the LMS platform,…
A vulnerability in the Border Gateway Protocol (BGP) enforce-first-as feature of Cisco Nexus 3000 Series Switches and Cisco Nexus 9000 Series Switches in standalone NX-OS mode could allow an unauthenticated, remote attacker to trigger BGP peer flaps, resulting in a denial of service (DoS) condition. This vulnerability is due to incorrect parsing of a transitive BGP attribute. An attacker could exploit this vulnerability by sending a crafted BGP update through an established BGP peer session. If…
A vulnerability in the access validation of internal REST APIs of Cisco Secure Workload could allow an unauthenticated, remote attacker to access site resources with the privileges of the Site Admin role. This vulnerability is due to insufficient validation and authentication when accessing REST API endpoints. An attacker could exploit this vulnerability if they are able to send a crafted API request to an affected endpoint. A successful exploit could allow the attacker to read sensitive inform…
A vulnerability in the SSL certificate handling of Cisco ThousandEyes Virtual Appliance could allow an authenticated, remote attacker to execute commands on the underlying operating system as the root user. This vulnerability is due to insufficient validation of user-supplied input. An authenticated attacker could exploit this vulnerability by uploading a crafted certificate to an affected device. A successful exploit could allow the attacker to execute arbitrary code as the root user on the un…
A vulnerability in the BrowserBot component of Cisco ThousandEyes Enterprise Agent could have allowed an authenticated, remote attacker to execute arbitrary commands on Agents on behalf of the BrowserBot synthetics orchestration process. Cisco has addressed this vulnerability in the Cisco ThousandEyes Enterprise Agent, and no customer action is needed. This vulnerability was due to insufficient input validation of command arguments that are supplied by the user. Prior to this vulnerability bein…
On April 23, 2026, the U.S. Cybersecurity and Infrastructure Security Agency (CISA) issued an update to V1: Emergency Directive (ED) 25-03: Identify and Mitigate Potential Compromise of Cisco Devices related to Cisco Secure Firewall Adaptive Security Appliance (ASA) and Cisco Secure Firewall Threat Defense (FTD) products. According to the update, the ArcaneDoor threat actor has developed a previously unknown persistence mechanism that is preserved across upgrading to the fixed releases that wer…
Written by: Austin Larsen, Tyler McLellan, Genevieve Stark, Dan Ebreo Introduction Google Threat Intelligence Group (GTIG) has continued to track an expansive extortion campaign by UNC6671, a threat actor operating under the "BlackFile" brand, that targets organizations via sophisticated voice phishing (vishing) and single sign-on (SSO) compromise. By leveraging adversary-in-the-middle (AiTM) techniques to bypass traditional perimeter defenses and multi-factor authentication (MFA), UNC6671 gain…
[Mise à jour du 11 juin 2026] Le 9 juin 2026, Microsoft a publié des versions correctives. [Publication initiale] Le 14 mai 2026, Microsoft a publié un avis de sécurité concernant la vulnérabilité CVE-2026-42897 affectant Exchange Server. Elle permet à un attaquant non authentifié de provoquer...
Multiple vulnerabilities in Cisco Catalyst SD-WAN Manager, formerly SD-WAN vManage, could allow a remote attacker to gain access to sensitive information, elevate privileges, or gain unauthorized access to the application. For more information about these vulnerabilities, see the Details section of this advisory. Cisco has released software updates that address these vulnerabilities. There are no workarounds that address these vulnerabilities. Cisco strongly recommends that customers upgrade to…
Following the initial publication of the Security Advisory about a denial of service (DoS) condition in Cisco Crosswork Network Controller and Cisco Network Services Orchestrator (NSO), additional information has been made available to the Cisco Product Security Incident Response Team (PSIRT). Upon further analysis, the Cisco PSIRT has reclassified this issue as a customer-configurable, resource management issue rather than a security vulnerability. This advisory is available at the following l…
Why Mythos (and other AI models) Make Continuous Exposure Visibility Critical For years, vulnerability discovery was naturally constrained by expertise, time, and scale. Finding meaningful security issues often required experienced researchers spending days or weeks understanding codebases, testing assumptions, reviewing implementations, and validating exploitability. That dynamic is changing rapidly. Recent developments around systems like Anthropic’s Project Glasswing 🔗 and Mythos, OpenAI’s …
Executive Summary Since our February 2026 report on AI-related threat activity, Google Threat Intelligence Group (GTIG) has continued to track a maturing transition from nascent AI-enabled operations to the industrial-scale application of generative models within adversarial workflows. This report, based on insights derived from Mandiant incident response engagements, Gemini, and GTIG’s proactive research, highlights the dual nature of the current threat environment where AI serves as both a so…
When a critical Linux kernel privilege escalation was publicly disclosed, Cloudflare's security and engineering teams detected, investigated, and mitigated the threat across our global fleet, confirming zero customer impact and no malicious exploitation.
Multiple vulnerabilities in Cisco Unity Connection could allow a remote attacker to execute arbitrary code on or conduct server-side request forgery (SSRF) attacks through an affected device. For more information about these vulnerabilities, see the Details section of this advisory. Cisco has released software updates that address these vulnerabilities. There are no workarounds that address these vulnerabilities. This advisory is available at the following link:https://sec.cloudapps.cisco.com/s…
Multiple vulnerabilities in Cisco Identity Services Engine (ISE) could allow a remote attacker to bypass authorization mechanisms or examine error messages to gain access to sensitive information on an affected device. For more information about these vulnerabilities, see the Details section of this advisory. Cisco has released software updates that address these vulnerabilities. There are no workarounds that address these vulnerabilities. This advisory is available at the following link:https:…
A vulnerability in the log file download functionality of Cisco Prime Infrastructure could allow an authenticated, remote attacker to download arbitrary log files from the server. This vulnerability is due to insufficient authorization checks on the download service API. An attacker could exploit this vulnerability by submitting a crafted URL request to an affected device. A successful exploit could allow the attacker to download sensitive log files that they would otherwise not have authorizat…
A vulnerability in the REST API of Cisco Slido could have allowed an authenticated, remote attacker to access the social profile data of other users or affect quiz and poll results. Cisco has addressed this vulnerability in Cisco Slido and no customer action is needed. This vulnerability existed because of the presence of an insecure direct object reference. Prior to this vulnerability being addressed, an attacker could have exploited this vulnerability by sending a crafted request to the vulne…
Multiple vulnerabilities in the web-based management interface of Cisco IoT Field Network Director Software could allow an authenticated, remote attacker to access files, execute commands, and cause denial of service (DoS) conditions on managed routers. For more information about these vulnerabilities, see the Details section of this advisory. Cisco has released software updates that address these vulnerabilities. There are no workarounds that address these vulnerabilities. This advisory is ava…
A vulnerability in the Simple Network Management Protocol (SNMP) subsystem of Cisco 350 Series Managed Switches (SG350) and Cisco 350X Series Stackable Managed Switches (SG350X) firmware could allow an authenticated, remote attacker to cause a denial of service (DoS) condition on an affected device. This vulnerability is due to improper error handling when parsing response data for a specific SNMP request. An attacker could exploit this vulnerability by sending a specific SNMP request to an aff…
On 6 May 2026, Palo Alto published a security advisory addressing a critical vulnerability affecting PAN-OS. This vulnerability allows an unauthenticated attacker to execute arbitrary code with root privileges. Palo Alto observed limited exploitation of this vulnerability. It is strongly recommended updating affected appliances as soon as patches will be available, and to apply workarounds and mitigation in the meantime.
On 29 April 2026, a high local privilege escalation vulnerability in the Linux kernel, tracked as CVE-2026-31431 and named "Copy Fail", was publicly disclosed. The vulnerability affects every mainstream Linux distributions shipping a kernel built since 2017. A public proof-of-concept exploit has been released. As of the date of this advisory, no distribution has shipped a fixed kernel package. The mainline fix was committed on 1 April 2026, but vendor updates are still pending across all major …
A breach claims the systems as well as the confidence that was, in retrospect, a major vulnerability
Introduction Advances in AI model-powered exploitation have demonstrated that general-purpose AI models can excel at vulnerability discovery, even without being purpose-built for the task. Eventually, capabilities such as these will be integrated directly into the development cycle, and code will be more difficult to exploit than ever; however, this transition creates a critical window of risk. As we harden existing software with AI, threat actors will use it to discover and exploit novel vulne…
Posted by Jiacheng Lu, Software Engineer, Google Pixel Team Google is continuously advancing the security of Pixel devices. We have been focusing on hardening the cellular baseband modem against exploitation. Recognizing the risks associated within the complex modem firmware, Pixel 9 shipped with mitigations against a range of memory-safety vulnerabilities. For Pixel 10, Google is advancing its proactive security measures further. Following our previous discussion on "Deploying Rust in Existing…
Posted by Adam Gavish, Google GenAI Security TeamIndirect prompt injection (IPI) is an evolving threat vector targeting users of complex AI applications with multiple data sources, such as Workspace with Gemini. This technique enables the attacker to influence the behavior of an LLM by injecting malicious instructions into the data or tools used by the LLM as it completes the user’s query. This may even be possible without any input directly from the user.IPI is not the kind of technical proble…
2025 marked a special year in the history of vulnerability rewards and bug bounty programs at Google: our 15th anniversary 🎉🎉🎉! Originally started in 2010, our vulnerabi…
Posted by Dirk Göhmann, Tony Mendez, and the Vulnerability Rewards Program Team2025 marked a special year in the history of vulnerability rewards and bug bounty programs at Google: our 15th anniversary 🎉🎉🎉! Originally started in 2010, our vulnerability reward program (VRP) has seen constant additions and expansions over the past decade and a half, clearly indicating the value the programs under this umbrella contribute to the safety and security of Google and its users, but also highlighting…
Le 15 octobre 2025, F5 a publié un avis de sécurité concernant entre autres la vulnérabilité CVE-2025-53521. Celle-ci affecte BIG-IP APM et permet à un attaquant non authentifié d'exécuter du code à distance. Le 29 mars 2026, l'éditeur indique que cette vulnérabilité est exploitée activement. Le...
We are opening our advanced Client-Side Security tools to all users, featuring a new cascading AI detection system. By combining graph neural networks and LLMs, we've reduced false positives by up to 200x while catching sophisticated zero-day exploits.
