In that article we mentioned the importance of encryption.
“With a browser password manager, someone with access to your browser could see your passwords in clear text, although Windows can be set to ask for authentication (the same you use at startup of your device).”
The typical behavior of browser password managers is to store passwords encrypted on disk, tied to your user account, and protected by the operating system.
But recently, a security researcher systematically tested every major Chromium-based browser for how they handle credentials in memory. The researcher found that Edge was the only one loading the entire password vault into plaintext process memory at startup, where it remains for the duration of the session.
Chrome and other Chromium browsers were observed to only decrypt a password when needed (autofill or “show password”), not the whole vault, and to use mechanisms like app‑bound encryption for keys. Edge does not use those protections in this context.
So, the researcher decided to write a proof-of-concept (PoC) demonstrating that accessing that vault doesn’t rely on zero-days or complex exploitation. It relies on the relatively simple ability to read process memory, which does require elevated privileges.
But when the researcher reported the issue to Microsoft, the response was underwhelming. The company’s official response was that the behavior is “by design.” The reasoning most likely is that this behavior speeds up sign‑in and autofill, and attackers would already need a compromised machine or elevated access to read RAM, which Microsoft treats as out of scope for this design decision.
Which is basically true. An attacker already needs significant foothold: for example, code execution on the box and the ability to read Edge’s process memory, often requiring elevated privileges. This is not a remote, unauthenticated bug in the browser, but the design makes post‑compromise credential harvesting easier. And it’s a capability many infostealers already have.
It’s just another thing an attacker can do once they’ve compromised your machine. Combined with this academic study from 2024, which found many password managers leak plaintext passwords into memory under some conditions, it leads us to repeat our advice.
Should you allow your browser to remember your passwords?
Your browser password manager gives you ease of use, but that costs you some security. Of course, password managers aren’t foolproof either, so it’s important to decide for yourself where you store your passwords.
If you’re confident the website is safe, and anyone that can access it under your account won’t learn anything new, feel free to store the password in your browser, but disable autofill so you stay in control.
Use MFA where possible. It enormously reduces the risk should someone get hold of your password. And refrain from using the browser password manager to store your credit card details or other sensitive personally identifiable information, such as medical information.
But we’d add that, among the major browsers, Edge appears to be the weakest option if you still choose to use a built‑in password manager.
Stop threats before they can do any harm.
Malwarebytes Browser Guard blocks phishing pages and malicious sites automatically. Free, one click to install. Add it to your browser →
Cybersecurity expert Tom Rønning finds Microsoft Edge loads all saved passwords into computer memory as cleartext, making them easy for hackers to steal.
The paradox of edge security describes how technologies designed to strengthen network defenses can also create new vulnerabilities. Edge devices improve performance and support localized threat detection by processing data closer to its source, yet modern enterprise environments often operate thousands of distributed endpoints. This rapid expansion of edge infrastructure increases the number of systems..
Over 130,000 users are at risk from fake TikTok downloader extensions on Chrome and Microsoft Edge. Researchers discovered these malicious tools use device fingerprinting to spy on users and steal sensitive browser data.
Have you ever been on a website when a pop-up suddenly asked for access to your camera, microphone, location, or notifications? Whether you clicked “allow,” dismissed it, or just wondered why it appeared, those permission requests aren’t always harmless. Some sites can abuse those permissions.
With Access Control, a new feature in Browser Guard, you decide exactly which websites can access your device and stop the rest. That means you choose which websites can:
Use your camera
Use your microphone
Access your location
Send you notifications
Further, not only can you control which websites have access to your devices, but you can also block websites or even require those specific sites to request permission every single time they try to gain access to your machines. You can always allow trusted sites to access your camera or location while blocking everything else.
Access Control is now available for Malwarebytes subscribers using Chrome and Edge browsers on a Windows device.
How to use Access Control
We designed Access Control to be both powerful and simple because we know every moment you spend getting set up is another moment you’re left unprotected.
How to use Access Control:
Install/Open Browser Guard: Click the Malwarebytes icon in your browser’s header
Access Dashboard: Click the Dashboard tab at the bottom of the extension panel.
Navigate to Access Control: On the left sidebar of the web page, select Access Control.
Manage Permissions: See visited websites, click “Allow” to enable or disable Malwarebytes’ ability to see visited sites.
Access Control requires some access to your browsing to protect you online
Access Control lets you choose individual sites to block and allow
This feature is rolling out in beta first, so you might see improvements and updates as we refine it. Currently, the feature works across Chrome and Edge, but will roll out to other browsers soon.
Access Control is another step toward making privacy simple and accessible. Not a subscriber yet? Check out Malwarebytes’ plans today to unlock this feature and more.
We don’t just report on threats—we remove them
Cybersecurity risks should never spread beyond a headline. Keep threats off your devices by downloading Malwarebytes today.
How modern age-verification laws, like the California Digital Age Assurance Act, dismantle the principle of data minimization by mandating the collection of sensitive personal data, effectively turning "don't know" into "must know" and knowledge into liability.
Por mais que você não saia procurando serviços de IA, eles acabam encontrando você de qualquer maneira. Todas as grandes empresas de tecnologia parecem sentir uma espécie de obrigação moral não apenas de desenvolver um assistente de IA, chatbot integrado ou agente autônomo, mas também de incorporá-lo aos seus produtos já consolidados e ativá-lo à força para dezenas de milhões de usuários. Aqui estão apenas alguns exemplos dos últimos seis meses:
O Google ativou o Gemini para todos os usuários do Chrome nos Estados Unidos, elevou ao máximo as funcionalidades do navegador, ampliou agressivamente o alcance dos Resumos de IA nos resultados de busca e incorporou um conjunto completo de recursos de IA nos seus serviços online (Gmail, Google Docs e outros).
A Apple integrou sua própria Apple Intelligence (convenientemente compartilhando a sigla AI) às versões mais recentes de seus sistemas operacionais em todos os tipos de dispositivos e na maioria de seus aplicativos nativos.
Por outro lado, entusiastas de tecnologia correram para criar seus próprios “Jarvis pessoais”, alugando instâncias de VPS ou acumulando Mac minis para executar o agente de IA OpenClaw. Infelizmente, os problemas de segurança do OpenClaw com as configurações padrão se mostraram tão graves que já foram considerados a maior ameaça de cibersegurança de 2026.
Além do incômodo de ter algo imposto à força, essa epidemia de IA traz riscos e dores de cabeça bem reais do ponto de vista prático. Assistentes de IA varrem e coletam todos os dados a que conseguem ter acesso, interpretando o contexto dos sites que você visita, analisando documentos salvos, lendo suas conversas e assim por diante. Isso dá às empresas de IA uma visão inédita e extremamente íntima da vida de cada usuário.
Um vazamento desses dados durante um ataque cibernético, seja a partir dos servidores do provedor de IA ou do cache armazenado na sua própria máquina, poderia ser catastrófico. Esses assistentes podem ver e armazenar em cache tudo o que você vê, inclusive dados normalmente protegidos por múltiplas camadas de segurança: informações bancárias, diagnósticos médicos, mensagens privadas e outras informações sensíveis. Analisamos em profundidade como isso pode acontecer quando examinamos os problemas do sistema Copilot+ Recall baseado em IA que a Microsoft também planejava impor a todos os usuários. Além disso, a IA pode consumir muitos recursos do sistema, utilizando RAM, ciclos de GPU e espaço de armazenamento, o que frequentemente resulta em uma queda perceptível no desempenho.
Para quem prefere ficar de fora dessa onda de IA e evitar esses assistentes baseados em redes neurais lançados às pressas e ainda imaturos, reunimos um guia rápido mostrando como desativar a IA em aplicativos e serviços populares.
Como desativar a IA no Google Docs, Gmail e Google Workspace
Os recursos de assistente de IA do Google no Gmail e no Google Docs são agrupados sob o termo “recursos inteligentes”. Além do modelo de linguagem de grande escala, esse conjunto inclui várias conveniências de menor importância, como adicionar automaticamente reuniões ao seu calendário quando você recebe um convite no Gmail. Infelizmente, trata-se de um pacote tudo ou nada: para se livrar da IA, é preciso desativar todos os “recursos inteligentes”.
Para fazer isso, abra o Gmail, clique no ícone Configurações (engrenagem) e selecione Ver todas as configurações. Na aba Geral, role até Recursos inteligentes do Google Workspace. Clique em Gerenciar as configurações de recursos inteligentes do Workspace e desative duas opções: Recursos inteligentes no Google Workspace e Recursos inteligentes em outros produtos do Google. Também recomendamos desmarcar a caixa ao lado de Ativar os recursos inteligentes no Gmail, Chat e Meet na mesma aba de configurações gerais. Depois disso, será necessário reiniciar os aplicativos do Google (o que normalmente ocorre de forma automática).
Como desativar os Resumos de IA na Pesquisa Google
É possível eliminar os Resumos de IA nos resultados da Pesquisa Google tanto em computadores quanto em smartphones (incluindo iPhones). A solução é a mesma em todos os dispositivos. A maneira mais simples de ignorar o resumo de IA caso a caso é adicionar -ia ao final da sua busca. Exemplo: como fazer uma pizza -ia. Infelizmente, esse método às vezes apresenta falhas, fazendo o Google afirmar abruptamente que não encontrou nenhum resultado para a sua consulta.
Se isso acontecer, você pode obter o mesmo resultado mudando o modo da página de resultados para Web. Nos resultados da pesquisa, localize os filtros logo abaixo da barra de busca e selecione Web. Caso não apareça imediatamente, procure essa opção dentro do botão Mais.
Uma solução mais radical é migrar para outro mecanismo de busca. Por exemplo, o DuckDuckGo não apenas rastreia menos os usuários e exibe poucos anúncios, como também oferece uma busca dedicada sem IA. Basta adicionar a página de pesquisa aos favoritos em noai.duckduckgo.com.
Como desativar recursos de IA no Chrome
Atualmente, o Chrome incorpora dois tipos de recursos de IA. O primeiro se comunica com os servidores do Google e é responsável por funções como o assistente inteligente, um agente autônomo de navegação e a busca inteligente. O segundo executa tarefas localmente, mais voltadas para utilidades, como identificar páginas de phishing ou agrupar abas do navegador. O primeiro grupo de configurações aparece com o rótulo AI mode, enquanto o segundo inclui o termo Gemini Nano.
Para desativar esses recursos, digite chrome://flags na barra de endereços do navegador e pressione Enter. Será exibida uma lista de flags do sistema, junto com uma barra de busca. Digite “AI” na barra de busca. Isso filtrará a longa lista para cerca de uma dúzia de recursos relacionados à IA (além de algumas outras configurações nas quais essas letras aparecem por coincidência dentro de palavras maiores). O segundo termo que você deve pesquisar nessa janela é “Gemini“.
Depois de revisar as opções, você pode desativar os recursos de IA indesejados ou simplesmente desativar todos. O mínimo recomendado inclui:
AI Mode Omnibox entrypoint
AI Entrypoint Disabled on User Input
Omnibox Allow AI Mode Matches
Prompt API for Gemini Nano
Prompt API for Gemini Nano with Multimodal Input
Defina todas essas opções como Disabled.
Como desativar recursos de IA no Firefox
Embora o Firefox não tenha chatbots integrados nem tenha (até agora) tentado impor recursos baseados em agentes aos usuários, o navegador inclui agrupamento inteligente de abas, uma barra lateral para chatbots e algumas outras funcionalidades. Em geral, a IA no Firefox é bem menos intrusiva do que no Chrome ou no Edge. Ainda assim, se você quiser desativá-la completamente, há duas maneiras de fazer isso.
O primeiro método está disponível nas versões mais recentes do Firefox. A partir da versão 148, uma seção dedicada chamada Controles de IA passou a aparecer nas configurações do navegador, embora as opções de controle ainda sejam um pouco limitadas. Você pode usar um único botão de alternância para Bloquear melhorias de IA, desativando completamente os recursos de IA. Você também pode especificar se deseja usar IA no próprio dispositivo (On-device AI), baixando pequenos modelos locais (atualmente apenas para traduções), e configurar provedores de chatbot de IA na barra lateral, escolhendo entre Anthropic Claude, ChatGPT, Copilot, Google Gemini e Le Chat Mistral.
O segundo caminho (para versões mais antigas do Firefox) exige acessar configurações ocultas do sistema. Digite about:config na barra de endereço, pressione Enter e clique no botão para confirmar que você aceita o risco de mexer nas configurações internas do navegador.
Uma extensa lista de configurações será exibida, juntamente com uma barra de busca. Digite “ML” para filtrar as opções relacionadas a machine learning.
Para desativar a IA no Firefox, alterne a configuração browser.ml.enabled para false. Isso deve desativar todos os recursos de IA de forma geral, mas fóruns da comunidade indicam que isso nem sempre é suficiente para resolver o problema. Para uma abordagem mais radical, defina os seguintes parâmetros como false (ou mantenha apenas aqueles de que você realmente precisa):
ml.chat.enabled
ml.linkPreview.enabled
ml.pageAssist.enabled
ml.smartAssist.enabled
ml.enabled
ai.control.translations
tabs.groups.smart.enabled
urlbar.quicksuggest.mlEnabled
Isso desativará integrações com chatbots, descrições de links geradas por IA, assistentes e extensões baseados em IA, tradução local de sites, agrupamento de abas e outros recursos baseados em IA.
Como desativar recursos de IA em aplicativos da Microsoft
A Microsoft conseguiu incorporar IA em praticamente todos os seus produtos, e desativá-la nem sempre é uma tarefa simples, especialmente porque, em alguns casos, a IA tem o hábito de reaparecer sozinha, sem qualquer ação do usuário.
Como desativar recursos de IA no Edge
O navegador da Microsoft está repleto de recursos de IA, que vão do Copilot à pesquisa automatizada. Para desativá-los, siga a mesma lógica usada no Chrome: digite edge://flags na barra de endereços do Edge, pressione Enter e, em seguida, digite “AI” ou “Copilot” na caixa de pesquisa. A partir daí, você pode desativar os recursos de IA indesejados, como:
Enable Compose (AI-writing) on the web
Edge Copilot Mode
Edge History AI
Outra maneira de se livrar do Copilot é digitar edge://settings/appearance/copilotAndSidebar na barra de endereço. Ali, você pode personalizar a aparência da barra lateral do Copilot e ajustar as opções de personalização para resultados e notificações. Não se esqueça de verificar também a seção Copilot em App-specific settings. Você encontrará alguns controles adicionais escondidos ali.
Como desativar o Microsoft Copilot
O Microsoft Copilot existe em duas versões: como um componente do Windows (Microsoft Copilot) e como parte do pacote Office (Microsoft 365 Copilot). As funções são semelhantes, mas você terá que desativar um ou ambos, dependendo exatamente do que os engenheiros de Redmond decidiram instalar na sua máquina.
A coisa mais simples que você pode fazer é desinstalar o aplicativo por completo. Clique com o botão direito na entrada Copilot no menu Iniciar e selecione Desinstalar. Se essa opção não estiver disponível, vá até a lista de aplicativos instalados (Iniciar → Configurações → Aplicativos) e desinstale o Copilot por lá.
Em determinadas versões do Windows 11, o Copilot está integrado diretamente ao sistema operacional, portanto uma simples desinstalação pode não funcionar. Nesse caso, você pode desativá-lo pelas configurações: Iniciar → Configurações → Personalização → Barra de Tarefas→ Desativar o Copilot.
Se você mudar de ideia no futuro, sempre poderá reinstalar o Copilot pela Microsoft Store.
Vale observar que muitos usuários reclamaram que o Copilot se reinstala automaticamente. Portanto, pode ser uma boa ideia fazer uma verificação semanal durante alguns meses para garantir que ele não tenha voltado. Para quem se sente confortável em mexer no Registro do Sistema (e entende as consequências disso), é possível seguir este guia detalhado para evitar o retorno silencioso do Copilot, desativando o parâmetro SilentInstalledAppsEnabled e adicionando/ativando o parâmetro TurnOffWindowsCopilot.
Como desativar o Microsoft Recall
O recurso Microsoft Recall, apresentado pela primeira vez em 2024, funciona tirando constantemente capturas de tela do seu computador e fazendo com que uma rede neural as analise. Todas essas informações extraídas são armazenadas em um banco de dados, que você pode pesquisar posteriormente usando um assistente de IA. Já escrevemos anteriormente, em detalhes, sobre os enormes riscos de segurança que o Microsoft Recall representa.
Sob pressão de especialistas em cibersegurança, a Microsoft foi obrigada a adiar o lançamento desse recurso de 2024 para 2025, reforçando significativamente a proteção dos dados armazenados. No entanto, o funcionamento básico do Recall permanece o mesmo: seu computador continua registrando cada movimento seu ao tirar capturas de tela constantemente e aplicar OCR ao conteúdo. E, embora o recurso não esteja mais ativado por padrão, vale absolutamente a pena verificar se ele não foi ativado na sua máquina.
Para verificar, vá até as configurações: Iniciar → Configurações → Privacidade e segurança →Recall e capturas de tela. Assegure-se de que a opção Salvar capturas de tela esteja desativada e clique em Excluir capturas de tela para limpar todos os dados coletados anteriormente, por precaução.
Como desativar a IA no Notepad e nas ações de contexto do Windows
A IA se infiltrou em praticamente todos os cantos do Windows, até mesmo no Explorador de Arquivos e no Notepad. Basta selecionar texto por engano em um aplicativo para que recursos de IA sejam acionados, o que a Microsoft chama de “Ações de IA”. Para desativar essa ação, vá para Iniciar → Configurações → Privacidade e segurança → Clique para executar.
O Notepad recebeu seu próprio tratamento com Copilot, portanto será necessário desativar a IA nele separadamente. Abra as configurações do Notepad, localize a seção Recursos de IA e desative o Copilot.
Por fim, a Microsoft também conseguiu incorporar o Copilot ao Paint. Infelizmente, até o momento não existe uma maneira oficial de desativar os recursos de IA dentro do próprio aplicativo Paint.
Como desativar a IA no WhatsApp
Em várias regiões, usuários do WhatsApp começaram a ver adições típicas de IA, como respostas sugeridas, resumos de mensagens gerados por IA e um novo botão Pergunte à Meta AI ou pesquise. Embora a Meta afirme que os dois primeiros recursos processam os dados localmente no dispositivo e não enviam suas conversas para os servidores da empresa, verificar isso não é tarefa simples. Felizmente, desativá-los é fácil.
Para desativar Sugestões de respostas, vá para Configurações → Conversas → Sugestões e respostas inteligentes e desative Sugestões de respostas. Você também pode desativar as Sugestões de figurinhas por IA nesse mesmo menu. Quanto aos resumos de mensagens gerados por IA, eles são gerenciados em outro local: Configurações → Notificações → Resumos de mensagens por IA.
Como desativar a IA no Android
Dada a grande variedade de fabricantes e versões do Android, não existe um manual único que sirva para todos os celulares. Hoje, vamos nos concentrar em eliminar os serviços de IA do Google, mas se você estiver usando um dispositivo da Samsung, Xiaomi ou outros, não se esqueça de verificar as configurações de IA do fabricante específico. Vale um aviso: eliminar completamente qualquer vestígio de IA pode ser uma tarefa difícil, se é que isso é realmente possível.
No Google Mensagens, os recursos de IA ficam nas configurações: toque na foto da sua conta, selecione Configurações do Mensagens, depois Gemini no app Mensagens e desative o assistente.
De modo geral, o chatbot Gemini funciona como um aplicativo independente que pode ser desinstalado acessando as configurações do telefone e selecionando Aplicativos. No entanto, como o plano do Google é substituir o tradicional Google Assistant pelo Gemini, desinstalá-lo pode se tornar difícil (ou até impossível) no futuro.
Se você não conseguir desinstalar completamente o Gemini, abra o aplicativo para desativar manualmente seus recursos. Toque no ícone do seu perfil, selecione Atividade dos apps do Gemini e escolha Desativar ou Desativar e excluir atividade. Em seguida, toque novamente no ícone do perfil e vá até a configuração Apps conectados (pode estar dentro da opção Inteligência pessoal). A partir daí, desative todos os aplicativos nos quais você não quer que o Gemini interfira.
Os recursos de IA no nível da plataforma da Apple, conhecidos coletivamente como Apple Intelligence, são relativamente simples de desativar. Nas configurações, tanto em desktops quanto em smartphones e tablets, basta procurar a seção Apple Intelligence e Siri. Aliás, dependendo da região e do idioma selecionado para o sistema operacional e para a Siri, o Apple Intelligence pode nem estar disponível para você ainda.
Outros artigos para ajudar você a ajustar as ferramentas de IA em seus dispositivos:
Throughout early 2026, SentinelOne’s® Digital Forensics & Incident Response (DFIR) team has responded to several incidents where FortiGate Next-Generation Firewall (NGFW) appliances have been compromised to establish a foothold into the targeted environment. Each incident was detected and stopped during the lateral movement phase of the attack.
Fortinet has disclosed and issued patches for several high-severity vulnerabilities allowing unauthorized access during the activity period of our investigations. Successful exploitation of these flaws allows an attacker to extract the configuration file from the FortiGate appliance, which frequently contains service account credentials and valuable network topology information for the targeted environment.
We observed a consistent theme: targeted organizations fail to retain sufficient logs on these appliances, which prevents understanding exactly how and when attackers gained access. The dwell time between initial perimeter device compromise to network compromise was drastically different across two incidents we investigated, ranging from 2 months to near instantaneous follow-on activity.
This post explores the actions that an attacker or attackers conducted following likely exploitation of two of these FortiGate appliances in different environments. It also provides defenders with guidance to investigate compromise of these appliances and subsequent infiltration activities.
FortiGate Appliance Compromise
FortiGate network appliances have considerable access to the environments they were installed to protect. In many configurations, this includes service accounts which are connected to the authentication infrastructure, such as Active Directory (AD) and Lightweight Directory Access Protocol (LDAP). This setup can enable the appliance to map roles to specific users by fetching attributes about the connection that’s being analyzed and correlating with the Directory information, which is useful in cases where role-based policies are set or for increasing response speed for network security alerts detected by the device.
However, such access is abused by actors who compromise FortiGate devices, as we have seen in these recent incidents.
Through December 2025 and February 2026, Fortinet products have reportedly been exploited via CVE-2025-59718 and CVE-2025-59719, two vulnerabilities impacting Fortinet products’ Single Sign On (SSO) mechanisms by failing to validate cryptographic signatures. In effect, this means an attacker who sends a crafted SSO token can achieve unauthenticated administrative access as the cryptographic signature is not verified.
Another vulnerability, CVE-2026-24858, was patched by Fortinet in late January: this vulnerability permitted attackers to log into FortiGate devices where FortiCloud SSO was enabled. Attackers exploited this flaw by logging into the victim’s device using the attacker’s FortiCloud account.
Once an attacker gains access in this way, they can run the command show full-configuration to extract the FortiGate device configuration file. Fortinet’s FortiOS devices, including FortiGate appliances, use a reversible form of encryption on the configuration files, meaning an attacker can then identify embedded service accounts and extract them.
However, other recent reports have detailed that actors are scanning for open instances and then attempting to log into FortiGate devices using common weak credentials, which means actors can access such devices without a weaponized exploit.
FortiGate Configs Abused to Enroll Rogue Domain Workstations
In one incident (IOCS: Incident 1), the compromise likely began in late November 2025 and remained undetected through February 2026. After accessing the appliance, the actor created a new local administrator account on the FortiGate device named support and used it to create 4 new firewall policies that allowed the account to traverse all zones (source: all; destination: all).
Activity then dropped to low volumes of traffic through some of these policies, suggesting the actor was periodically checking that access was still available before later shifting to noisier network activity.
This pattern is consistent with an initial access broker (IAB) establishing a foothold and then selling it on to another actor. Insufficient FortiGate log retention meant we could only reconstruct the activity window rather than identify the precise initial access vector.
In February 2026, an attacker likely extracted the configuration file, which contained encrypted service account LDAP credentials. Evidence demonstrates the attacker authenticated to the AD using clear text credentials from the fortidcagent service account, suggesting the attacker decrypted the configuration file and extracted the service account credentials.
The service account was then used to authenticate to the victim’s environment from IP address 193.24.211[.]61. The attacker used the mS-DS-MachineAccountQuota attribute to join two rogue workstations to the AD; by default, this setting permits a standard account to join up to 10 workstations to the domain.
Joining the attacker’s workstation to the AD provided the attacker with more access to the environment with fewer security controls. The rogue workstation names are:
WIN-X8WRBOSK0OF
WIN-YRSXLEONJY2
Per Validin’s records on IP address 193.24.211[.]61, the IP has consistently shown an RDP port open with an exposed Windows system with workstation ID WIN-1J7L3SQSTMS. We did not see this workstation ID during our incident, but given the consistent nature of this system on the IP address, this workstation ID should be considered suspect.
The actor then performed network scanning across the environment, which generated security alerts and prevented further lateral movement. Identity logs showed massive volumes of failed logins indicating password spraying attempts, which originated from the FortiGate appliance IP address. There were also multiple delete.me file artifacts that suggest the actor likely used the SoftPerfect Network Scanner for enumeration.
There were multiple failed login attempts during the period of heavy activity in February, including from IP addresses 185.156.73[.]62 and 185.242.246[.]127, which are registered to networks in Ukraine and Kazakhstan, respectively.
FortiGate Access Exploited to Deploy RMM Tools and Steal NTDS
In another case we investigated in late January (IOCS: Incident 2), a threat actor accessed the organization’s FortiGate appliance and created a local administrator account with the name ssl-admin. As in the previous investigation, it is likely that the actor again extracted the configuration from the FortiGate appliance and decrypted it to harvest the AD administrator credentials.
Within 10 minutes of creating a local account on the FortiGate appliance, the actor logged into several servers in the victim’s environment with the built-in Domain Administrator account. Server authentication logs confirmed a series of successful Network (Type 3) and Remote Interactive (Type 10/RDP) logins originating from the FortiGate VPN-assigned IP range.
On one of the servers, the attacker launched SQL Server Management Studio (SSMS) but did not connect to any databases, possibly searching for stored connection details or credentials in the application.
The actor began staging files in the system’s C:\ProgramData\USOShared directory, a technique we have observed across multiple incidents. The attacker downloaded two Remote Monitoring and Management (RMM) tools: Pulseway and MeshAgent, which are legitimate system administration tools that are frequently abused by threat actors to achieve a deeper foothold in the target environment.
The actor abused legitimate cloud storage functionality by hosting the Pulseway application on a Google Cloud Storage URL at hxxps://storage.googleapis[.]com/apply-main/windows_agent_x64[.]msi, which is likely attacker-controlled. The MeshAgent RMM was installed on the domain controller and a file share. The actor modified a Windows Registry value SystemComponent=1 to hide MeshAgent from the “Programs and Features” list.
These tools were used to create Windows Scheduled Tasks named JavaMainUpdate and MeshUserTask for Pulseway and MeshAgent, respectively. The actor also downloaded malware from a cloud storage bucket via PowerShell from Amazon Web Services (AWS) Simple Storage Service (S3) hostname fastdlvrss[.]s3[.]us-east-1[.]amazonaws[.]com. Like the Google Cloud Storage URL, this is a legitimate Amazon resource that was likely registered by the attacker for unauthorized purposes. This stage used the following command:
The attacker gave the malicious DLLs the same names as legitimate Java files, causing the application to load the malware via DLL side-loading instead of the real components. The Java-loaded payload beaconed to two domains: ndibstersoft[.]com and neremedysoft[.]com. These payloads were executed against other servers on the network using PsExec, including the primary and secondary domain controllers.
The actor then created a Volume Shadow Copy backup of the primary domain controller via Windows Management Instrumentation Controls (WMIC) and extracted the NTDS.dit file and SYSTEM registry hive from the backup using the makecab command, then used makecab to compress each file.
The malicious Java application then established a connection on port 443 to IP address 172.67.196[.]232, a Cloudflare-owned IP address with thousands of domain records. The connection was terminated after 8 minutes, and the compressed NTDS.dit and registry hive files were deleted. This sequence of events suggests the attacker uploaded the data to their infrastructure during this session.
Following this activity, we observed no evidence of the threat actor leveraging new or additional user accounts. While the actor may have attempted to crack passwords from the data, no such credential usage was identified between the time of credential harvesting and incident containment.
Conclusion
NGFW appliances have become ubiquitous because they provide strong network monitoring capabilities for organizations by integrating security controls of a firewall with other management features, such as AD. However, these devices are high-value targets for actors with a variety of motivations and skill levels, from state-aligned actors conducting espionage to financially motivated attacks such as ransomware.
As Amazon Security recently wrote, lower skilled threat actors have been boosted by integrating large language models (LLM) into their workflows, making attacks easier and more automated despite demonstrating limited knowledge after exploitation. SentinelOne does not see indications that the campaigns identified in the aforementioned cases are associated with the threat actor tracked by Amazon Security, particularly given the relatively high dwell time between initial access and further activity mentioned in the first incident.
However, organizations should prepare for increases in attack volume at network edge devices as attackers find novel ways to bypass LLM safeguards: these appliances are valuable targets and are often exposed to the open internet. LLMs are often trained on these products and can readily supply information to actors that facilitates gaining access and understanding how to navigate from network appliances deeper into the targeted environment without the knowledge uplift previously required by threat actor crews.
Organizations should consider that FortiGate and other edge devices typically do not permit security software to be installed on the appliance, such as endpoint detection and response (EDR) tools. The best defense for these appliances is to apply strong administrative access controls and to keep the software patched to prevent exploitation. Further, both of these investigations were hindered by insufficient FortiGate log retention. Organizations should ensure they have at least 14 days of log retention on NGFW appliances like FortiGate, though 60-90 days is much better when possible.
SentinelOne recommends sending all logs to a Security Incident & Event Monitoring (SIEM) or similar log aggregation system, as attackers may delete logs from local systems after establishing access, but they can’t delete what’s already been sent to a security center. A SIEM can help at each stage of an attack:
UEBA and Identity Intelligence: User entity and behavior analytics (UEBA) within a SIEM baselines “normal behavior for every administrator, allowing the SIEM to immediately flag a login if it originates from an unrecognized device or “impossible travel” location. By identifying these anomalies at the moment of entry, the SIEM can alert defenders even if the attacker is using perfectly valid, stolen credentials.
Detection of Configuration Access and Credential Extraction: A SIEM monitors the FortiGate audit logs for sensitive CLI commands like show full-configuration or backup exports occurring outside of maintenance windows. By correlating these actions with an unusual admin login location, the SIEM can alert security teams that a configuration file and the credentials within it may have been compromised.
Spotting Unauthorized Account Creation: When a threat actor creates a local “backdoor” admin account, the SIEM immediately flags the user-creation event as a high-priority anomaly. It compares this new account against a “whitelist” of authorized admins, ensuring that any account created without a linked Change Management ticket is treated as an active breach.
Monitoring Malware Downloads & C2 Traffic: SIEMs analyze network flow data to identify internal systems reaching out to known malicious IPs or suspicious domains via PowerShell. By detecting the “heartbeat” of a C2 channel, the SIEM allows defenders to kill the connection before the attacker can begin exfiltrating data.
Preserving Evidence Against Log Deletion: Because the FortiGate appliance streams its logs to the SIEM in real-time, the attacker cannot hide their tracks by clearing the local logs. The SIEM maintains an immutable record of every command the attacker typed, providing the forensic evidence needed to understand the full scope of the intrusion.
Neutralizing the Threat with Automation: Modern SIEMs now come with automation built in allowing the SIEM to trigger automated playbooks the millisecond an attack is detected, such as instantly disabling the compromised service account or “shunning” the attacker’s IP at the network perimeter. By removing the need for human intervention in the initial response, automation slashes the attacker’s dwell time, effectively neutralizing the breach before malware can begin to spread.
Below, we share guidance for organizations and other incident responders on how to investigate a suspected FortiGate intrusion of this nature.
Forensic Investigation Guidance
FortiGate
Malicious SSO/Unexpected Logins:
Search system logs for Log ID 0100032001 (Admin login successful, method sso).
Check for usernames matching public IOCs (e.g., cloud-init@mail.io, cloud-noc@mail.io).
Configuration Download:
Search for Log ID 0100032095 (System config file has been downloaded).
The timestamp confirms when the attacker exported the configuration file.
Malicious Local Admin Account Creation:
Identify the exact creation time and source IP using Log ID 0100044547 (Object attribute configured) with cfgpath="user.local" or cfgpath="system.admin".
VPN Sessions:
Identify the attacker’s source IP by analyzing the remip field in VPN tunnel logs.
Look for Log ID 0101039424 (SSL VPN tunnel up) or 0101037138 (IPsec tunnel up).
Note the internal IP addresses for later correlation with evidence from the Domain Controller(s).
Domain Controllers
Rogue Computer Account Creation (Domain Join):
Windows Event ID 4741: Confirms the creation of a computer account.
Verify the Subject: Security ID matches FortiGate’s stolen LDAP bind account.
Use the SubjectLogonId from 4741 to find the corresponding Event ID 4624 (Logon Type 3) on the same domain controller to extract the Source Network Address (should match the attacker’s internal VPN IP identified via FortiGate logs).
Directory Service Changes (Advanced Audit):
If enabled, Event ID 5136 shows exact attributes set during the join (e.g., missing SPNs, modified UAC values), indicating the use of automated tools like Impacket.
DNS Record Creation:
DNS Server Audit log (Event ID 515 – Record Create): Records the creation of the host (A) record, including workstation name and timestamp.
Microsoft-Windows-DNSServer/Analytical log: Provides the Client IP (internal VPN IP) and the QNAME (workstation name being registered).
Active Directory
Find Malicious Computer Objects:
Check mS-DS-CreatorSID: If multiple rogue computers share the same SID, and it belongs to the Fortinet LDAP service account, compromise is confirmed.
Check for defined SPNs: The lack of SPNs is a red flag and a likely malicious indicator.
Note whenCreated: This attribute stores the exact time the object was added.
Originating Domain Controller and Time:
Check the sAMAccountName attribute to identify the Originating DSA (the domain controller that originally created the object) and the Originating Time.
ssl-admin – Incident 2, FortiGate local administrative account
support – Incident 1, FortiGate local administrative account
Windows Workstation Names
WIN-1J7L3SQSTMS – Incident 1, Windows hostname of RDP service hosted on attacker IP 193.24.211[.]61
WIN-X8WRBOSK0OF – Incident 1, rogue workstation ID
WIN-YRSXLEONJY2 – Incident 1, rogue workstation ID
Third-Party Trademark Disclaimer:
All third-party product names, logos, and brands mentioned in this publication are the property of their respective owners and are for identification purposes only. Use of these names, logos, and brands does not imply affiliation, endorsement, sponsorship, or association with the third-party.
I’m a big advocate of password managers. Granted, there are better alternatives for passwords like passkeys, but if a provider offers nothing but password options, which many do, you can’t do much about that. So, for the time being we seem to be stuck with passwords.
Every reputable password manager claims that they can’t see your passwords, even if they wanted to. But researchers have found that these “zero‑knowledge” cloud password managers are more vulnerable than their marketing suggests.
The researchers also warn that this is not an immediate cause for panic. For a full‑on password leakage to happen, would require rare, high‑end failures such as a malicious or fully compromised server combined with specific design weaknesses and features being enabled.
The underlying “problem” is that most of these password managers are cloud-based. Very handy if you’re working on another device and need access, but it also enlarges the attack surface. Sharing your passwords with another device or another user opens it up to the possibility of unwanted access.
The researchers tested a number of different vendors, including LastPass, Bitwarden, and Dashlane, and devised several attack scenarios that could allow the recovery of passwords.
Weaknesses
Password managers with groups of users
In groups, the sharing of recovery keys, group keys, and admin public keys often means they are fetched from the server without an authenticity guarantee. Meaning that under the right circumstances, an attacker could gain access.
When a group admin has enabled policies such as “auto or manual recovery,” it’s possible to silently change them using a compromised server if there’s no integrity protection on the org “policy blob” (a small configuration file).
Weak encryption on compromised server
Your password manager takes your master password and runs it through PBKDF2 many times (e.g., 600,000 rounds) before storing a hash. But on a compromised server, an attacker could turn down the number of iterations to, say, 2, which makes the master password easy to guess or brute-force.
Account recovery options
On a compromised server an attacker could change the policy blob and change the settings to “auto recovery” and send it to the clients. Switching to auto‑recovery helps the attacker because it lets the system hand over your vault keys without anyone having to click “approve” or even notice it happening.
So, the attacker can turn what should be a rare, user‑visible emergency process into a silent, routine mechanism they can abuse to pull out vault keys at scale or in a stealthy, targeted way.
Backwardscompatibility
To avoid locking out users on old clients, providers keep supporting deprecated key hierarchies and non‑AEAD (Authenticated Encryption with Associated Data) modes such as CBC (Cipher Block Chaining) without robust integrity checks. This opens the door to classic downgrade attacks where the server coaxes a client into using weaker schemes and then gradually recovers plaintext.
How to stay safe
We want to emphasize that these attacks would be very targeted and require a high level of compromise. So, cloud password managers are still much safer than password reuse and spreadsheets, but their “zero‑knowledge” claims don’t hold up under nation state‑level type of attacks.
After responsible disclosure, many of the issues have already been patched or mitigated, reducing the number of possible attacks.
Many of the demonstrated attacks require specific enterprise‑style features (account recovery, shared vaults, org membership) or older/legacy clients to be in use. So be extra careful with those.
Enable multi-factor authentication for important accounts, so the attacker will not have enough by just obtaining your password.
We don’t just report on threats—we help safeguard your entire digital identity
Cybersecurity risks should never spread beyond a headline. Protect your, and your family’s, personal information by using identity protection.
Over the past few years, we’ve been observing and monitoring the espionage activities of HoneyMyte (aka Mustang Panda or Bronze President) within Asia and Europe, with the Southeast Asia region being the most affected. The primary targets of most of the group’s campaigns were government entities.
As an APT group, HoneyMyte uses a variety of sophisticated tools to achieve its goals. These tools include ToneShell, PlugX, Qreverse and CoolClient backdoors, Tonedisk and SnakeDisk USB worms, among others. In 2025, we observed HoneyMyte updating its toolset by enhancing the CoolClient backdoor with new features, deploying several variants of a browser login data stealer, and using multiple scripts designed for data theft and reconnaissance.
An early version of the CoolClient backdoor was first discovered by Sophos in 2022, and TrendMicro later documented an updated version in 2023. Fast forward to our recent investigations, we found that CoolClient has evolved quite a bit, and the developers have added several new features to the backdoor. This updated version has been observed in multiple campaigns across Myanmar, Mongolia, Malaysia and Russia where it was often deployed as a secondary backdoor in addition to PlugX and LuminousMoth infections.
In our observations, CoolClient was typically delivered alongside encrypted loader files containing encrypted configuration data, shellcode, and in-memory next-stage DLL modules. These modules relied on DLL sideloading as their primary execution method, which required a legitimate signed executable to load a malicious DLL. Between 2021 and 2025, the threat actor abused signed binaries from various software products, including BitDefender, VLC Media Player, Ulead PhotoImpact, and several Sangfor solutions.
Variants of CoolClient abusing different software for DLL sideloading (2021–2025)
The latest CoolClient version analyzed in this article abuses legitimate software developed by Sangfor. Below, you can find an overview of how it operates. It is worth noting that its behavior remains consistent across all variants, except for differences in the final-stage features.
Overview of CoolClient execution flow
However, it is worth noting that in another recent campaign involving this malware in Pakistan and Myanmar, we observed that HoneyMyte has introduced a newer variant of CoolClient that drops and executes a previously unseen rootkit. A separate report will be published in the future that covers the technical analysis and findings related to this CoolClient variant and the associated rootkit.
CoolClient functionalities
In terms of functionality, CoolClient collects detailed system and user information. This includes the computer name, operating system version, total physical memory (RAM), network details (MAC and IP addresses), logged-in user information, and descriptions and versions of loaded driver modules. Furthermore, both old and new variants of CoolClient support file upload to the C2, file deletion, keylogging, TCP tunneling, reverse proxy listening, and plugin staging/execution for running additional in-memory modules. These features are still present in the latest versions, alongside newly added functionalities.
In this latest variant, CoolClient relies on several important files to function properly:
Filename
Description
Sang.exe
Legitimate Sangfor application abused for DLL sideloading.
libngs.dll
Malicious DLL used to decrypt loader.dat and execute shellcode.
loader.dat
Encrypted file containing shellcode and a second-stage DLL. Parameter checker and process injection activity reside here.
time.dat
Encrypted configuration file.
main.dat
Encrypted file containing shellcode and a third-stage DLL. The core functionality resides here.
Parameter modes in second-stage DLL
CoolClient typically requires three parameters to function properly. These parameters determine which actions the malware is supposed to perform. The following parameters are supported.
Parameter
Actions
No parameter
· CoolClient will launch a new process of itself with the install parameter. For example: Sang.exe install.
install
CoolClient decrypts time.dat.
Adds new key to the Run registry for persistence mechanism.
Creates a process named write.exe.
Decrypts and injects loader.dat into a newly created write.exe process.
Checks for service control manager (SCM) access.
Checks for multiple AV processes such as 360sd.exe, zhudongfangyu.exe and 360desktopservice64.exe.
Installs a service named media_updaten and starts it.
If the current user is in the Administrator group, creates a new process of itself with the passuac parameter to bypass UAC.
work
Creates a process named write.exe.
Decrypts and injects loader.dat into a newly spawned write.exe process.
passuac
Bypasses UAC and performs privilege elevation.
Checks if the machine runs Windows 10 or a later version.
Impersonates svchost.exe process by spoofing PEB information.
Creates a scheduled task named ComboxResetTask for persistence. The task executes the malware with the work parameter.
Elevates privileges to admin by duplicating an access token from an existing elevated process.
Final stage DLL
The write.exe process decrypts and launches the main.dat file, which contains the third (final) stage DLL. CoolClient’s core features are implemented in this DLL. When launched, it first checks whether the keylogger, clipboard stealer, and HTTP proxy credential sniffer are enabled. If they are, CoolClient creates a new thread for each specific functionality. It is worth noting that the clipboard stealer and HTTP proxy credential sniffer are new features that weren’t present in older versions.
Clipboard and active windows monitor
A new feature introduced in CoolClient is clipboard monitoring, which leverages functions that are typically abused by clipboard stealers, such as GetClipboardData and GetWindowTextW, to capture clipboard information.
CoolClient also retrieves the window title, process ID and current timestamp of the user’s active window using the GetWindowTextW API. This information enables the attackers to monitor user behavior, identify which applications are in use, and determine the context of data copied at a given moment.
The clipboard contents and active window information are encrypted using a simple XOR operation with the byte key 0xAC, and then written to a file located at C:\ProgramData\AppxProvisioning.xml.
HTTP proxy credential sniffer
Another notable new functionality is CoolClient’s ability to extract HTTP proxy credentials from the host’s HTTP traffic packets. To do so, the malware creates dedicated threads to intercept and parse raw network traffic on each local IP address. Once it is able to intercept and parse the traffic, CoolClient starts extracting proxy authentication credentials from HTTP traffic intercepted by the malware’s packet sniffer.
The function operates by analyzing the raw TCP payload to locate the Proxy-Connection header and ensure the packet is relevant. It then looks for the Proxy-Authorization: Basic header, extracts and decodes the Base64-encoded credential and saves it in memory to be sent later to the C2.
Function used to find and extract Base64-encoded credentials from HTTP proxy-authorization headers
C2 command handler
The latest CoolClient variant uses TCP as the main C2 communication protocol by default, but it also has the option to use UDP, similar to the previous variant. Each incoming payload begins with a four-byte magic value to identify the command family. However, if the command is related to downloading and running a plugin, this value is absent. If the client receives a packet without a recognized magic value, it switches to plugin mode (mechanism used to receive and execute plugin modules in memory) for command processing.
Magic value
Command category
CC BB AA FF
Beaconing, status update, configuration.
CD BB AA FF
Operational commands such as tunnelling, keylogging and file operations.
No magic value
Receive and execute plugin module in memory.
0xFFAABBCC – Beacon and configuration commands
Below is the command menu to manage client status and beaconing:
Command ID
Action
0x0
Send beacon connection
0x1
Update beacon timestamp
0x2
Enumerate active user sessions
0x3
Handle incoming C2 command
0xFFAABBCD – Operational commands
This command group implements functionalities such as data theft, proxy setup, and file manipulation. The following is a breakdown of known subcommands:
Command ID
Action
0x0
Set up reverse tunnel connection
0x1
Send data through tunnel
0x2
Close tunnel connection
0x3
Set up reverse proxy
0x4
Shut down a specific socket
0x6
List files in a directory
0x7
Delete file
0x8
Set up keylogger
0x9
Terminate keylogger thread
0xA
Get clipboard data
0xB
Install clipboard and active windows monitor
0xC
Turn off clipboard and active windows monitor
0xD
Read and send file
0xE
Delete file
CoolClient plugins
CoolClient supports multiple plugins, each dedicated to a specific functionality. Our recent findings indicate that the HoneyMyte group actively used CoolClient in campaigns targeting Mongolia, where the attackers pushed and executed a plugin named FileMgrS.dll through the C2 channel for file management operations.
Further sample hunting in our telemetry revealed two additional plugins: one providing remote shell capability (RemoteShellS.dll), and another focused on service management (ServiceMgrS.dll).
ServiceMgrS.dll – Service management plugin
This plugin is used to manage services on the victim host. It can enumerate all services, create new services, and even delete existing ones. The following table lists the command IDs and their respective actions.
Command ID
Action
0x0
Enumerate services
0x1 / 0x4
Start or resume service
0x2
Stop service
0x3
Pause service
0x5
Create service
0x6
Delete service
0x7
Set service to start automatically at boot
0x8
Set service to be launched manually
0x9
Set service to disabled
FileMgrS.dll – File management plugin
A few basic file operations are already supported in the operational commands of the main CoolClient implant, such as listing directory contents and deleting files. However, the dedicated file management plugin provides a full set of file management capabilities.
Command ID
Action
0x0
List drives and network resources
0x1
List files in folder
0x2
Delete file or folder
0x3
Create new folder
0x4
Move file
0x5
Read file
0x6
Write data to file
0x7
Compress file or folder into ZIP archive
0x8
Execute file
0x9
Download and execute file using certutil
0xA
Search for file
0xB
Send search result
0xC
Map network drive
0xD
Set chunk size for file transfers
0xF
Bulk copy or move
0x10
Get file metadata
0x11
Set file metadata
RemoteShellS.dll – Remote shell plugin
Based on our analysis of the main implant, the C2 command handler did not implement remote shell functionality. Instead, CoolClient relied on a dedicated plugin to enable this capability. This plugin spawns a hidden cmd.exe process, redirecting standard input and output through pipes, which allows the attacker to send commands into the process and capture the resulting output. This output is then forwarded back to the C2 server for remote interaction.
CoolClient plugin that spawns cmd.exe with redirected I/O and forwards command output to C2
Browser login data stealer
While investigating suspicious ToneShell backdoor traffic originating from a host in Thailand, we discovered that the HoneyMyte threat actor had downloaded and executed a malware sample intended to extract saved login credentials from the Chrome browser as part of their post-exploitation activities. We will refer to this sample as Variant A. On the same day, the actor executed a separate malware sample (Variant B) targeting credentials stored in the Microsoft Edge browser. Both samples can be considered part of the same malware family.
During a separate threat hunting operation focused on HoneyMyte’s QReverse backdoor, we retrieved another variant of a Chrome credential parser (Variant C) that exhibited significant code similarities to the sample used in the aforementioned ToneShell campaign.
The malware was observed in countries such as Myanmar, Malaysia, and Thailand, with a particular focus on the government sector.
The following table shows the variants of this browser credential stealer employed by HoneyMyte.
Variant
Targeted browser(s)
Execution method
MD5 hash
A
Chrome
Direct execution (PE32)
1A5A9C013CE1B65ABC75D809A25D36A7
B
Edge
Direct execution (PE32)
E1B7EF0F3AC0A0A64F86E220F362B149
C
Chromium-based browsers
DLL side-loading
DA6F89F15094FD3F74BA186954BE6B05
These stealers may be part of a new malware toolset used by HoneyMyte during post-exploitation activities.
Initial infection
As part of post-exploitation activity involving the ToneShell backdoor, the threat actor initially executed the Variant A stealer, which targeted Chrome credentials. However, we were unable to determine the exact delivery mechanism used to deploy it.
A few minutes later, the threat actor executed a command to download and run the Variant B stealer from a remote server. This variant specifically targeted Microsoft Edge credentials.
Within the same hour that Variant B was downloaded and executed, we observed the threat actor issue another command to exfiltrate the Firefox browser cookie file (cookies.sqlite) to Google Drive using a curl command.
Unlike Variants A and B, which use hardcoded file paths, the Variant C stealer accepts two runtime arguments: file paths to the browser’s Login Data and Local State files. This provides greater flexibility and enables the stealer to target any Chromium-based browser such as Chrome, Edge, Brave, or Opera, regardless of the user profile or installation path. An example command used to execute Variant C is as follows:
In this context, the Login Data file is an SQLite database that stores saved website login credentials, including usernames and AES-encrypted passwords. The Local State file is a JSON-formatted configuration file containing browser metadata, with the most important value being encrypted_key, a Base64-encoded AES key. It is required to decrypt the passwords stored in the Login Data database and is also encrypted.
When executed, the malware copies the Login Data file to the user’s temporary directory as chromeTmp.
Function that copies Chrome browser login data into a temporary file (chromeTmp) for exfiltration
To retrieve saved credentials, the malware executes the following SQL query on the copied database:
SELECT origin_url, username_value, password_value FROM logins
This query returns the login URL, stored username, and encrypted password for each saved entry.
Next, the malware reads the Local State file to extract the browser’s encrypted master key. This key is protected using the Windows Data Protection API (DPAPI), ensuring that the encrypted data can only be decrypted by the same Windows user account that created it. The malware then uses the CryptUnprotectData API to decrypt this key, enabling it to access and decrypt password entries from the Login Data SQLite database.
With the decrypted AES key in memory, the malware proceeds to decrypt each saved password and reconstructs complete login records.
Finally, it saves the results to the text file C:\Users\Public\Libraries\License.txt.
Login data stealer’s attribution
Our investigation indicated that the malware was consistently used in the ToneShell backdoor campaign, which was attributed to the HoneyMyte APT group.
Another factor supporting our attribution is that the browser credential stealer appeared to be linked to the LuminousMoth APT group, which has previously been connected to HoneyMyte. Our analysis of LuminousMoth’s cookie stealer revealed several code-level similarities with HoneyMyte’s credential stealer. For example, both malware families used the same method to copy targeted files, such as Login Data and Cookies, into a temporary folder named ChromeTmp, indicating possible tool reuse or a shared codebase.
Code similarity between HoneyMyte’s saved login data stealer and LuminousMoth’s cookie stealer
Both stealers followed the same steps: they checked if the original Login Data file existed, located the temporary folder, and copied the browser data into a file with the same name.
Based on these findings, we assess with high confidence that HoneyMyte is behind this browser credential stealer, which also has a strong connection to the LuminousMoth APT group.
Document theft and system information reconnaissance scripts
In several espionage campaigns, HoneyMyte used a number of scripts to gather system information, conduct document theft activities and steal browser login data. One of these scripts is a batch file named 1.bat.
1.bat – System enumeration and data exfiltration batch script
The script starts by downloading curl.exe and rar.exe into the public folder. These are the tools used for file transfer and compression.
Batch script that downloads curl.exe and rar.exe from HoneyMyte infrastructure and executes them for file transfer and compression
It then collects network details and downloads and runs the nbtscan tool for internal network scanning.
Batch script that performs network enumeration and saves the results to the log.dat file for later exfiltration
During enumeration, the script also collects information such as stored credentials, the result of the systeminfo command, registry keys, the startup folder list, the list of files and folders, and antivirus information into a file named log.dat. It then uploads this file via FTP to http://113.23.212[.]15/pub/.
Batch script that collects registry, startup items, directories, and antivirus information for system profiling
Next, it deletes both log.dat and the nbtscan executable to remove traces. The script then terminates browser processes, compresses browser-related folders, retrieves FileZilla configuration files, archives documents from all drives with rar.exe, and uploads the collected data to the same server.
Finally, it deletes any remaining artifacts to cover its tracks.
Ttraazcs32.ps1 – PowerShell-based collection and exfiltration
The second script observed in HoneyMyte operations is a PowerShell file named Ttraazcs32.ps1.
Similar to the batch file, this script downloads curl.exe and rar.exe into the public folder to handle file transfers and compression. It collects computer and user information, as well as network details such as the public IP address and Wi-Fi network data.
All gathered information is written to a file, compressed into a password-protected RAR archive and uploaded via FTP.
In addition to system profiling, the script searches multiple drives including C:\Users\Desktop, Downloads, and drives D: to Z: for recently modified documents. Targeted file types include .doc, .xls, .pdf, .tif, and .txt, specifically those changed within the last 60 days. These files are also compressed into a password-protected RAR archive and exfiltrated to the same FTP server.
t.ps1 – Saved login data collection and exfiltration
The third script attributed to HoneyMyte is a PowerShell file named t.ps1.
The script requires a number as a parameter and creates a working directory under D:\temp with that number as the directory name. The number is not related to any identifier. It is simply a numeric label that is probably used to organize stolen data by victim. If the D drive doesn’t exist on the victim’s machine, the new folder will be created in the current working directory.
The script then searches the system for Chrome and Chromium-based browser files such as Login Data and Local State. It copies these files into the target directory and extracts the encrypted_key value from the Local State file. It then uses Windows DPAPI (System.Security.Cryptography.ProtectedData) to decrypt this key and writes the decrypted Base64-encoded key into a new file named Local State-journal in the same directory. For example, if the original file is C:\Users\$username \AppData\Local\Google\Chrome\User Data\Local State, the script creates a new file C:\Users\$username\AppData\Local\Google\Chrome\User Data\Local State-journal, which the attacker can later use to access stored credentials.
PowerShell script that extracts and decrypts the Chrome encrypted_key from the Local State file before writing the result to a Local State-journal file
Once the credential data is ready, the script verifies that both rar.exe and curl.exe are available. If they are not present, it downloads them directly from Google Drive. The script then compresses the collected data into a password-protected archive (the password is “PIXELDRAIN”) and uploads it to pixeldrain.com using the service’s API, authenticated with a hardcoded token. Pixeldrain is a public file-sharing service that attackers abuse for data exfiltration.
Script that compresses data with RAR, and exfiltrates it to Pixeldrain via API
This approach highlights HoneyMyte’s shift toward using public file-sharing services to covertly exfiltrate sensitive data, especially browser login credentials.
Conclusion
Recent findings indicate that HoneyMyte continues to operate actively in the wild, deploying an updated toolset that includes the CoolClient backdoor, a browser login data stealer, and various document theft scripts.
With capabilities such as keylogging, clipboard monitoring, proxy credential theft, document exfiltration, browser credential harvesting, and large-scale file theft, HoneyMyte’s campaigns appear to go far beyond traditional espionage goals like document theft and persistence. These tools indicate a shift toward the active surveillance of user activity that includes capturing keystrokes, collecting clipboard data, and harvesting proxy credential.
Organizations should remain highly vigilant against the deployment of HoneyMyte’s toolset, including the CoolClient backdoor, as well as related malware families such as PlugX, ToneShell, Qreverse, and LuminousMoth. These operations are part of a sophisticated threat actor strategy designed to maintain persistent access to compromised systems while conducting high-value surveillance activities.
Email remains the main means of business correspondence at organizations. It can be set up either using on-premises infrastructure (for example, by deploying Microsoft Exchange Server) or through cloud mail services such as Microsoft 365 or Gmail. However, some organizations do not provide domain-level access to their cloud email. As a result, attackers who have compromised the domain do not automatically gain access to email correspondence and must resort to additional techniques to read it.
This research describes how ToddyCat APT evolved its methods to gain covert access to the business correspondence of employees at target companies. In the first part, we review the incidents that occurred in the second half of 2024 and early 2025. In the second part of the report, we focus in detail on how the attackers implemented a new attack vector as a result of their efforts. This attack enables the adversary to leverage the user’s browser to obtain OAuth 2.0 authorization tokens. These tokens can then be utilized outside the perimeter of the compromised infrastructure to access corporate email.
In a previous post on the ToddyCat group, we described the TomBerBil family of tools, which are designed to extract cookies and saved passwords from browsers on user hosts. These tools were written in C# and C++.
Yet, analysis of incidents from May to June 2024 revealed a new variant implemented in PowerShell. It retained the core malicious functionality of the previous samples but employed a different implementation approach and incorporated new commands.
A key feature of this version is that it was executed on domain controllers on behalf of a privileged user, accessing browser files via shared network resources using the SMB protocol.
Besides supporting the Chrome and Edge browsers, the new version also added processing for Firefox browser files.
The tool was launched using a scheduled task that executed the following command line:
The script begins by creating a new local directory, which is specified in the $baseDir variable. The tool saves all data it collects into this directory.
The script defines a function named parseFile, which accepts the full file path as a parameter. It opens the C:\programdata\uhosts.txt file and reads its content line by line using .NET Framework classes, returning the result as a string array. This is how the script forms an array of host names.
For each host in the array, the script attempts to establish an SMB connection to the shared resource c$, constructing the path in the \\\c$\users\ format. If the connection is successful, the tool retrieves a list of user directories present on the remote host. If at least one directory is found, a separate folder is created for that host within the $baseDir working directory:
In the next stage, the script iterates through the user folders discovered on the remote host, skipping any folders specified in the $filter_users variable, which is defined upon launching the tool. For the remaining folders, three directories are created in the script’s working folder for collecting data from Google Chrome, Mozilla Firefox, and Microsoft Edge.
Next, the tool uses the default account to search for the following Chrome and Edge browser files on the remote host:
Login Data: a database file that contains the user’s saved logins and passwords for websites in an encrypted format
Local State: a JSON file containing the encryption key used to encrypt stored data
Cookies: a database file that stores HTTP cookies for all websites visited by the user
History: a database that stores the browser’s history
These files are copied via SMB to the local folder within the corresponding user and browser folder hierarchy. Below is a code snippet that copies the Login Data file:
The same procedure is applied to Firefox files, with the tool additionally traversing through all the user profile folders of the browser. Instead of the files described above for Chrome and Edge, the script searches for files which have names from the $firefox_files array that contain similar information. The requested files are also copied to the tool’s local folder.
The copied files are encrypted using the Data Protection API (DPAPI). The previous version of TomBerBil ran on the host and copied the user’s token. As a result, in the user’s current session DPAPI was used to decrypt the master key, and subsequently, the files. The updated server-side version of TomBerBil copies files containing the user encryption keys that are used by DPAPI. These keys, combined with the user’s SID and password, grant the attackers the ability to decrypt all the copied files locally.
With TomBerBil, the attackers automatically collected user cookies, browsing history, and saved passwords, while simultaneously copying the encryption keys needed to decrypt the browser files. The connection to the victim’s remote hosts was established via the SMB protocol, which significantly complicated the detection of the tool’s activity.
TomBerBil in PowerShell
As a rule, such tools are deployed at later stages, after the adversary has established persistence within the organization’s internal infrastructure and obtained privileged access.
Detection
To detect the implementation of this attack, it’s necessary to set up auditing for access to browser folders and to monitor network protocol connection attempts to those folders.
title: Access To Sensitive Browser Files Via Smb
id: 9ac86f68-9c01-4c9d-897a-4709256c4c7b
status: experimental
description: Detects remote access attempts to browser files containing sensitive information
author: Kaspersky
date: 2025-08-11
tags:
- attack.credential-access
- attack.t1555.003
logsource:
product: windows
service: security
detection:
event:
EventID: '5145'
chromium_files:
ShareLocalPath|endswith:
- '\User Data\Default\History'
- '\User Data\Default\Network\Cookies'
- '\User Data\Default\Login Data'
- '\User Data\Local State'
firefox_path:
ShareLocalPath|contains: '\AppData\Roaming\Mozilla\Firefox\Profiles'
firefox_files:
ShareLocalPath|endswith:
- 'key3.db'
- 'signons.sqlite'
- 'key4.db'
- 'logins.json'
condition: event and (chromium_files or firefox_path and firefox_files)
falsepositives: Legitimate activity
level: medium
In addition, auditing for access to the folders storing the DPAPI encryption key files is also required.
title: Access To System Master Keys Via Smb
id: ba712364-cb99-4eac-a012-7fc86d040a4a
status: experimental
description: Detects remote access attempts to the Protect file, which stores DPAPI master keys
references:
- https://www.synacktiv.com/en/publications/windows-secrets-extraction-a-summary
author: Kaspersky
date: 2025-08-11
tags:
- attack.credential-access
- attack.t1555
logsource:
product: windows
service: security
detection:
selection:
EventID: '5145'
ShareLocalPath|contains: 'windows\System32\Microsoft\Protect'
condition: selection
falsepositives: Legitimate activity
level: medium
Stealing emails from Outlook
The modified TomBerBil tool family proved ineffective at evading monitoring tools, compelling the threat actor to seek alternative methods for accessing the organization’s critical data. We discovered an attempt to gain access to corporate correspondence files in the local Outlook storage.
The Outlook application stores OST (Offline Storage Table) files for offline use. The names of these files contain the address of the mailbox being cached. Outlook uses OST files to store a local copy of data synchronized with mail servers: Microsoft Exchange, Microsoft 365, or Outlook.com. This capability allows users to work with emails, calendars, contacts, and other data offline, then synchronize changes with the server once the connection is restored.
However, access to an OST file is blocked by the application while Outlook is running. To copy the file, the attackers created a specialized tool called TCSectorCopy.
TCSectorCopy
This tool is designed for block-by-block copying of files that may be inaccessible by applications or the operating system, such as files that are locked while in use.
The tool is a 32-bit PE file written in C++. After launch, it processes parameters passed via the command line: the path to the source file to be copied and the path where the result should be saved. The tool then validates that the source path is not identical to the destination path.
Validating the TCSectorCopy command line parameters
Next, the tool gathers information about the disk hosting the file to be copied: it determines the cluster size, file system type, and other parameters necessary for low-level reading.
Determining the disk’s file system type
TCSectorCopy then opens the disk as a device in read-only mode and sequentially copies the file content block by block, bypassing the standard Windows API. This allows the tool to copy even the files that are locked by the system or other applications.
The adversary uploaded this tool to target host and used it to copy user OST files:
Having obtained the OST files, the attackers processed them using a separate tool to extract the email correspondence content.
XstReader
XstReader is an open-source C# tool for viewing and exporting the content of Microsoft Outlook OST and PST files. The attackers used XstReader to export the content of the previously copied OST files.
XstReader is executed with the -e parameter and the path to the copied file. The -e parameter specifies the export of all messages and their attachments to the current folder in the HTML, RTF, and TXT formats.
XstExport.exe -e <email>@<domain>.ost2
After exporting the data from the OST file, the attackers review the list of obtained files, collect those of interest into an archive, and exfiltrate it.
Stealing data with TCSectorCopy and XstReader
Detection
To detect unauthorized access to Outlook OST files, it’s necessary to set up auditing for the %LOCALAPPDATA%\Microsoft\Outlook\ folder and monitor access events for files with the .ost extension. The Outlook process and other processes legitimately using this file must be excluded from the audit.
title: Access To Outlook Ost Files
id: 2e6c1918-08ef-4494-be45-0c7bce755dfc
status: experimental
description: Detects access to the Outlook Offline Storage Table (OST) file
author: Kaspersky
date: 2025-08-11
tags:
- attack.collection
- attack.t1114.001
logsource:
product: windows
service: security
detection:
event:
EventID: 4663
outlook_path:
ObjectName|contains: '\AppData\Local\Microsoft\Outlook\'
ost_file:
ObjectName|endswith: '.ost'
condition: event and outlook_path and ost_file
falsepositives: Legitimate activity
level: low
The TCSectorCopy tool accesses the OST file via the disk device, so to detect it, it’s important to monitor events such as Event ID 9 (RawAccessRead) in Sysmon. These events indicate reading directly from the disk, bypassing the file system.
As we mentioned earlier, TCSectorCopy receives the path to the OST file via a command line. Consequently, detecting this tool’s malicious activity requires monitoring for a specific OST file naming pattern: the @ symbol and the .ost extension in the file name.
Example of detecting TCSectorCopy activity in KATA
Stealing access tokens from Outlook
Since active file collection actions on a host are easily tracked using monitoring systems, the attackers’ next step was gaining access to email outside the hosts where monitoring was being performed. Some target organizations used the Microsoft 365 cloud office suite. The attackers attempted to obtain the access token that resides in the memory of processes utilizing this cloud service.
In the OAuth 2.0 protocol, which Microsoft 365 uses for authorization, the access token is used when requesting resources from the server. In Outlook, it is specified in API requests to the cloud service to retrieve emails along with attachments. Its disadvantage is its relatively short lifespan; however, this can be enough to retrieve all emails from a mailbox while bypassing monitoring tools.
The access token is stored using the JWT (JSON Web Tokens) standard. The token content is encoded using Base64. JWT headers for Microsoft applications always specify the typ parameter with the JWT value first. This means that the first 18 characters of the encoded token will always be the same.
The attackers used SharpTokenFinder to obtain the access token from the user’s Outlook application. This tool is written in C# and designed to search for an access token in processes associated with the Microsoft 365 suite. After launch, the tool searches the system for the following processes:
“TEAMS”
“WINWORD”
“ONENOTE”
“POWERPNT”
“OUTLOOK”
“EXCEL”
“ONEDRIVE”
“SHAREPOINT”
If these processes are found, the tool attempts to open each process’s object using the OpenProcess function and dump their memory. To do this, the tool imports the MiniDumpWriteDump function from the dbghelp.dll file, which writes user mode minidump information to the specified file. The dump files are saved in the dump folder, located in the current SharpTokenFinder directory. After creating dump files for the processes, the tool searches for the following string pattern in each of them:
"eyJ0eX[a-zA-Z0-9\\._\\-]+"
This template uses the first six symbols of the encoded JWT token, which are always the same. Its structures are separated by dots. This is sufficient to find the necessary string in the process memory dump.
Example of a JWT Token
In the incident being described, the local security tools (EPP) blocked the attempt to create the OUTLOOK.exe process dump using SharpTokenFinder, so the operator used ProcDump from the Sysinternals suite for this purpose:
procdump64.exe -accepteula -ma OUTLOOK.exe
dir c:\windows\temp\OUTLOOK.EXE_<id>.dmp
c:\progra~1\winrar\rar.exe a -k -r -s -m5 -v100M %temp%\dmp.rar c:\windows\temp\OUTLOOK.EXE_<id>.dmp
Here, the operator executed ProcDump with the following parameters:
accepteula silently accepts the license agreement without displaying the agreement window.
ma indicates that a full process dump should be created.
exe is the name of the process to be dumped.
The dir command is then executed as a check to confirm that the file was created and is not zero size. Following this validation, the file is added to a dmp.rar archive using WinRAR. The attackers sent this file to their host via SMB.
Detection
To detect this technique, it’s necessary to monitor the ProcDump process command line for names belonging to Microsoft 365 application processes.
title: Dump Of Office 365 Processes Using Procdump
id: 5ce97d80-c943-4ac7-8caf-92bb99e90e90
status: experimental
description: Detects Office 365 process names in the command line of the procdump tool
author: kaspersky
date: 2025-08-11
tags:
- attack.lateral-movement
- attack.defense-evasion
- attack.t1550.001
logsource:
category: process_creation
product: windows
detection:
selection:
Product: 'ProcDump'
CommandLine|contains:
- 'teams'
- 'winword'
- 'onenote'
- 'powerpnt'
- 'outlook'
- 'excel'
- 'onedrive'
- 'sharepoint'
condition: selection
falsepositives: Legitimate activity
level: high
Below is an example of the ProcDump tool from the Sysinternals package used to dump the Outlook process memory, detected by Kaspersky Anti Targeted Attack (KATA).
Example of Outlook process dump detection in KATA
Takeaways
The incidents reviewed in this article show that ToddyCat APT is constantly evolving its techniques and seeking new ways to conceal its activity aimed at gaining access to corporate correspondence within compromised infrastructure. Most of the techniques described here can be successfully detected. For timely identification of these techniques, we recommend using both host-based EPP solutions, such as Kaspersky Endpoint Security for Business, and complex threat monitoring systems, such as Kaspersky Anti Targeted Attack. For comprehensive, up-to-date information on threats and corresponding detection rules, we recommend Kaspersky Threat Intelligence.
Entrar
