Huntress observed attackers abusing a compromised Microsoft 365 OAuth credential for ConnectWise Cloud Backup, formerly SkyKick, to sign in to 173 partner and customer tenants. The credential works against every tenant the app is consented in, letting attackers obtain Microsoft Graph tokens and reach mailboxes, files, and directory data. In at least one tenant, a ConnectWise app added itself to the Global Administrator role and created new client secrets. Token acquisition came from a single IP on October 15, moving through tenants in sorted order. ConnectWise closed Cloud Backup access until all credentials are reset, and issued tokens stay valid for up to 90 minutes.
A malicious Arch User Repository account uploaded typosquatted packages, including yyay, which imitates the popular yay helper, with an install scriptlet that runs arbitrary downloaded code as root at pacman install or upgrade time. The payload was pulled from a personal project on a Russian code-hosting site, and the attacker left the normal install functions empty while placing the malicious command outside them. A malware scanner flagged yyay with high confidence, and a maintainer confirmed the payloads, banned the account, and deleted its packages. The incident follows an earlier AUR malware wave in July that briefly halted uploads across the repository.
Bitdefender detailed Midnight Mimosa, a campaign in which low-cost Android phones built on MediaTek platforms ship with a preinstalled system component, com.android.system.lite, that runs with system privileges. It grants itself accessibility, notification, and SMS permissions, installs and removes apps silently, performs ad fraud through an invisible overlay, and turns devices into residential proxy relays and botnet nodes. It disables the Play Store while dropping payloads to evade Play Protect. Bitdefender also found 13 related Play Store apps, and most infections are in Mexico, France, and Italy. Because the component sits in the system partition, users cannot remove it without firmware-level intervention.
Ledger confirmed that a hardware wallet sold through the reseller CryptoBilis contained an unauthorized hardware implant, following reports of stolen funds from buyers in Southeast Asia. A pseudonymous blockchain investigator put losses above 86 million dollars, while Ledger described a limited number of affected devices, all sold through that reseller. Ledger paused sales through CryptoBilis, which halted all hardware wallet sales pending the investigation. Ledger advised buyers from the past 90 days not to set up unconfigured devices, and to move assets to a new signer with a fresh seed if a device was already initialized. Users should avoid unauthorized resellers.
Attackers hijacked refrigeration controllers running older Danfoss control software at several French supermarket chains, including Intermarche, Carrefour Market, Monoprix, and Leclerc, abusing a built-in defrost feature to raise temperatures and spoil stock. French outlets reported 250 to 350 affected stores, significant financial losses, and food waste, and the Paris prosecutor opened an investigation. Researchers linked two flaws in the Danfoss AK-SM 8xxA platform, CVE-2025-41450 for authentication bypass and CVE-2025-41451 for command injection, though how the controllers were reached is unconfirmed. One researcher found 2,752 Danfoss systems exposed online, 1,590 of them in France.
Socket, StepSecurity, and GitGuardian reported a resurgence of the GhostAction supply chain campaign, in which more than 500 GitHub accounts committed a credential-stealing workflow to tens of thousands of repositories since October 7. The attacker likely used maintainer personal access tokens leaked in infostealer logs, starting with the author of the pyxel game engine and spreading to other accounts. The injected workflow, disguised as a security audit, scans the working tree and full git history against 13 secret patterns, including AWS, Anthropic, OpenAI, and GitLab keys, and exfiltrates them over plain HTTP to a hardcoded address. Earlier GhostAction activity hit 772 repositories and thousands of secrets.
iVerify detailed P7 DarkSword, a new variant of the commercial DarkSword iOS exploit kit that has circulated among financially motivated and other actors since late 2025. It reduces its on-device footprint and adds on-device keychain and cryptocurrency wallet theft plus two-way command-and-control, running inside SpringBoard and polling for commands every 15 seconds. Collected data includes iCloud Keychain, Apple Notes, Photos, and wallet contents, with commands for disk, wallet, and photo scanning and file transfer. The chain exploits WebKit and Core Audio flaws, CVE-2025-24201 and CVE-2025-31200, patched in earlier iOS releases, and is distributed through fake websites, poisoned analytics domains, and bogus crypto platforms.
Socket and StepSecurity reported that version 0.5.144 of the tensorlake npm package, a TypeScript SDK, was published with obfuscated malware tied to the Shai-Hulud supply chain attack. A preinstall hook launches a loader that runs a worm under the Bun runtime. It harvests npm, GitHub, AWS, Vault, Kubernetes, and SSH credentials, .env files, cryptocurrency wallets, and configuration for Claude, Cursor, and other tools, then enumerates the victim's other packages and republishes them with forged provenance to spread. A PowerShell monitor watches the stolen GitHub token and runs a destructive handler if the victim revokes it, so remove the package before rotating that token.
Google disclosed that attackers hijacked the .gh, .sl, and .as country-code registries and changed authoritative DNS records to pass domain-validation checks, obtaining at least twelve TLS certificates for seven Google-owned names including google.com.gh, google.sl, and google.as. Let's Encrypt issued eleven and ZeroSSL one, all logged between September 22 and 27 and since revoked, and Chrome also blocked them through CRLSets. Such certificates let an attacker impersonate the real sites, though Google has not confirmed any misuse and says its own systems were not breached. Google advises domain owners to monitor certificate transparency logs and publish strict CAA records tied to their certificate authority.
The FBI and Secret Service warned that FortiBleed, a credential-harvesting campaign against internet-facing FortiGate firewalls and SSL VPN gateways, remains active. The Russian-speaking operation uses credential stuffing and password spraying, then deploys a Go tool called FortigateSniffer to intercept authentication traffic, exploiting reused or leaked credentials and legacy SHA-256 password storage rather than a software flaw. It had collected more than 86,644 working device credentials across 194 countries as of June. Attackers create new admin accounts and reuse session cookies for persistence, and some victims are locked out when original accounts are changed or deleted. Operator overlaps link it to INC and Lynx ransomware.