Last updated: September 29, 2026 at 8:42 AM UTC
All 891 Vulnerability 357 Breach 144 Threat 383 Defense 7

Microsoft says fake HR compliance emails fooled 35,000 people across 26 countries - phishing kit captured login tokens even with MFA enabled

Microsoft disclosed Monday that a phishing campaign between April 14 and 16 hit 35,000+ users across 13,000+ organizations in 26 countries (92% in the US). Lures impersonated internal HR with subjects like 'Internal case log issued under conduct policy.' Each email had a PDF attachment with a 'Review Case Materials' link that walked victims through Cloudflare CAPTCHAs and a final adversary-in-the-middle (AiTM) Microsoft sign-in page. AiTM proxies the real Microsoft login and captures session tokens after MFA - so traditional MFA is bypassed. Healthcare (19%), financial services (18%), and professional services (11%) were the most-targeted sectors.

Check
Search Exchange Online logs for emails between April 14-16 with subjects containing 'conduct policy' or 'awareness case log.' Hunt sign-in logs for OAuth grants from acceptable-use-policy-calendly.de or compliance-protectionoutlook.de.
Affected
Microsoft 365 / Entra ID tenants with users on traditional MFA (push, SMS, TOTP). AiTM bypasses any non-phishing-resistant MFA factor - only FIDO2 hardware keys and Windows Hello are immune. US users in healthcare, life sciences, financial services, and professional services are at acute risk based on Microsoft's targeting data.
Fix
Migrate users to phishing-resistant MFA (FIDO2 hardware keys, Windows Hello, passkeys) for all accounts. Enable Conditional Access policies that require token binding for high-privilege accounts. Turn on Zero-hour auto purge in Defender for Office 365 to retroactively quarantine campaign emails. Revoke session tokens for any user who visited a fake sign-in page.

China-linked spies breached the IBM subsidiary that runs IT for Italian government agencies and critical industries

La Repubblica reported a significant breach at Sistemi Informativi, a wholly-owned IBM Italy subsidiary that manages IT infrastructure for Italian public agencies and key industries. Multiple intelligence sources attribute the attack to Salt Typhoon, the China-linked espionage group that has hit US telecoms (AT&T, Verizon, Viasat), Canadian telecom firms, the US Army National Guard, Dutch government networks, and now Italian critical infrastructure. Salt Typhoon's hallmark is patience - prolonged data exfiltration, silent network observation, and infrastructure compromise rather than fast theft. The group has been active since at least 2019 and has reportedly hit 200+ companies across 80 countries.

Check
If your organization uses managed IT services for critical infrastructure (utilities, transport, healthcare, government), audit your provider's separation between corporate IT and customer environments this week.
Affected
Italian government agencies and key industries using Sistemi Informativi for IT infrastructure. More broadly: any organization where a single integrator holds access to multiple government databases - the breach pattern lets Salt Typhoon map critical infrastructure across many victims through one compromise. European telecoms and managed service providers are at acute risk.
Fix
Demand from any managed IT provider written attestation that customer environments are network-segregated from their corporate IT. Hunt for Salt Typhoon indicators: unauthorized configuration changes on edge devices, traffic to known Demodex C2 infrastructure, and anomalous data flows to Asian hosting providers. Treat the Italian breach as a reason to escalate vendor security reviews this quarter.

Scammers used Telegram's built-in mini-apps to impersonate Apple, NVIDIA, and Disney for crypto fraud and Android malware - all running on the same backend

CTM360 disclosed a large-scale fraud platform called FEMITBOT that uses Telegram's Mini App feature to host crypto scams, impersonate major brands, and distribute Android malware. The platform impersonates Apple, Coca-Cola, Disney, eBay, IBM, NVIDIA, BBC, and others - all backed by the same shared infrastructure identified by a common API response. The mini-apps display fake balances, countdown timers, and limited-time offers inside Telegram's WebView. Some campaigns push fake Android APKs hosted on the same domain as the API to ensure valid TLS certificates. Meta and TikTok tracking pixels measure conversion rates.

Check
Brief staff that any Telegram bot promoting cryptocurrency investments, asking them to deposit funds, or prompting them to install an APK is fraud - regardless of which brand the bot claims to represent.
Affected
Telegram users worldwide who interact with bots claiming to represent major brands. Acute risk for cryptocurrency-curious users targeted by 'investment opportunity' lures, and for Android users sideloading APKs from Telegram-shared links. Organizations whose brand is being impersonated face customer-trust damage even though the breach is in user behavior, not company systems.
Fix
Block sideloading of APKs on managed Android devices and require Google Play Protect to remain enabled. For brand protection teams: monitor Telegram for bots using your company name and report via Telegram's official channels - though the platform's Mini App vetting is essentially nonexistent so reactive moderation is the only path. Treat any 'official' Telegram bot as unverified by default.

Hackers are mass-encrypting websites by exploiting last week's cPanel flaw - 44,000 servers compromised so far in 'Sorry' ransomware attacks

Update on the cPanel flaw covered April 30: attackers are now mass-exploiting CVE-2026-41940 to deploy a Linux ransomware called 'Sorry' that encrypts websites and demands payment to unlock them. Shadowserver confirms at least 44,000 cPanel hosts have been compromised, with hundreds of victim sites already showing up in Google search results. The Sorry encryptor is written in Go, uses ChaCha20 with an embedded RSA-2048 public key (so victims cannot recover files without the attacker's private key), and appends '.sorry' to filenames. KnownHost reports the cPanel flaw was being exploited as a zero-day since at least February 23.

Check
If you run any cPanel or WHM server and have not yet patched, treat the server as already compromised - patch immediately, then start incident response rather than just resuming operations.
Affected
All cPanel and WHM versions before the April 28 emergency patch. ~1.5 million internet-exposed cPanel instances per Shodan, with 44,000 confirmed compromised. Hosting providers, web agencies, e-commerce sites on shared hosting, and any small business website on cPanel are in scope. Anyone whose cPanel was internet-reachable between February 23 and April 28 should assume compromise even if they patched promptly.
Fix
Patch cPanel to a fixed version. After patching, hunt for indicators of compromise (Sorry's '.sorry' file extension, unfamiliar admin sessions, cron entries pointing to /tmp/, modified /var/cpanel/sessions/raw/ files). Restore from clean backups predating February 23 if possible. Block cPanel ports (2082-2087, 2095-2096) at the firewall to non-trusted IPs. Rotate every credential the cPanel host had access to.

New 'ConsentFix v3' attack lets criminals take over Microsoft 365 accounts even when MFA and passkeys are turned on

Push Security disclosed ConsentFix v3, a new attack that lets criminals take over Microsoft 365 accounts even if the victim has MFA and phishing-resistant passkeys turned on. The trick: instead of stealing a password, the attacker tricks the user into pasting a Microsoft authorization URL into a phishing page during what looks like a routine login. That URL contains a one-time code that the attacker exchanges for permanent access tokens. v3 automates the whole attack with Cloudflare Pages phishing sites, Pipedream webhook automation, and tenant fingerprinting that customizes the lure to each target organization's branding.

Check
Brief any Microsoft 365 admin or developer that any 'verification step' that asks them to paste a URL containing 'localhost' into a webpage is hostile, no matter how legitimate the page looks.
Affected
Any Microsoft 365 / Entra ID tenant. The attack bypasses MFA, passkeys, and most Conditional Access policies by abusing pre-consented Microsoft first-party apps. Acute risk for organizations whose admins, developers, or DevOps engineers regularly use Azure CLI - those users won't suspect a fake Azure CLI authorization page. Cloudflare Pages and Pipedream both look legitimate in network telemetry.
Fix
Apply token binding to trusted devices and require Conditional Access for first-party Microsoft apps where possible. Hunt Azure sign-in logs for Azure CLI authentications from unfamiliar IPs, especially against accounts that don't normally use it. Train developers to verify out-of-band any 'verification step' that asks them to paste URLs into a webpage. Use app authentication restrictions to limit which first-party apps can issue refresh tokens.

Attackers poisoned 60+ Ruby gems and Go modules, then waited for CI pipelines to install them and steal credentials

Socket disclosed a fresh wave of supply-chain attacks targeting Ruby gems and Go modules: more than 60 typosquatted packages were uploaded to RubyGems and the Go module registry, designed to look like legitimate dependencies developers might pull into a CI pipeline. Once installed, the packages exfiltrate environment variables (which typically include AWS keys, GitHub tokens, and database credentials in CI environments) to attacker-controlled servers. The targeting is deliberate: typosquats picked names close to popular gems and Go libraries. This is the same operational pattern as the SAP npm compromise covered Wednesday, but targeting Ruby and Go ecosystems.

Check
Review your CI pipelines for any Ruby gem or Go module added in the past month, and confirm every package name matches the canonical upstream exactly.
Affected
Any organization running CI/CD pipelines that install Ruby gems or Go modules without strict pinning. Particularly acute for organizations with broad CI environment variables (AWS_SECRET_ACCESS_KEY, GITHUB_TOKEN, DATABASE_URL exposed to install scripts). Developer workstations are also exposed when developers run 'gem install' or 'go get' without verifying package names.
Fix
Pin every Ruby gem and Go module to specific versions and verify the upstream name matches. Move CI secrets out of environment variables and into ephemeral credential providers (OIDC for AWS, GitHub's masked secrets, Hashicorp Vault). Review CI logs for installs of packages whose names look like typosquats. Use Socket, Snyk, or equivalent tools to flag suspicious packages before install.

Vietnamese fraudsters used Google's no-code app platform to send Facebook phishing emails that passed every spam check, then sold the stolen accounts back to victims

Guardio documented a Vietnamese-linked fraud operation that has stolen roughly 30,000 Facebook business accounts by abusing Google's AppSheet no-code platform as a phishing relay. Because the phishing emails come from noreply@appsheet.com (a real Google address), they pass SPF, DKIM, and DMARC checks that normally catch fake-Meta emails. The lures impersonate Meta Support and threaten account deletion within 24 hours unless the user 'submits an appeal.' Stolen credentials, 2FA codes, and government ID photos are exfiltrated to Telegram. The operators then sell the stolen accounts back to victims through their own recovery service.

Check
Brief every staff member who manages a Facebook business account that any email from 'noreply@appsheet.com' claiming to be Meta is hostile, regardless of how legitimate the formatting looks.
Affected
Facebook Business account owners worldwide, with 68.6% of victims based in the US. Acute risk for marketing teams, social media managers, and small business owners who manage Facebook ad accounts. Any organization using the same Facebook business account for paid ads since 2024 is in the broader target pool. Stolen accounts often hold credit card data and ad spend history.
Fix
Block emails from noreply@appsheet.com unless your organization legitimately uses Google AppSheet. Train staff that real Meta support never asks for 2FA codes via email. Enable Meta Business Manager 2FA with hardware keys (not SMS). For organizations already compromised, contact Meta Business Help directly through facebook.com - the 'recovery service' is the same operation that took the account.

Cyber spies are quietly stealing engineering blueprints and GPS data from Russian aviation companies

Kaspersky disclosed a previously undocumented cyber-espionage group called HeartlessSoul that has been targeting Russian government agencies and aviation companies since at least September 2025 to steal geographic information system (GIS) data - the specialized files containing detailed maps of roads, engineering networks, terrain, and strategic facilities. The targeting suggests state-aligned interest in Russian infrastructure mapping rather than financial gain. Kaspersky did not name a likely sponsor but the targeting profile is consistent with a Ukraine-aligned or Western-aligned operator. The group uses tailored phishing, custom malware, and persistent network access.

Check
If your organization handles GIS data for any government or critical infrastructure customer, assume your sector is now an active target and tighten access controls on map data this week.
Affected
Russian government agencies and aviation companies are the named targets, but the technique is generic: any organization holding detailed GIS files for critical infrastructure (electric grid, telecoms, water, road, rail, military bases) is in the broader target pool. Engineering and architecture firms working on infrastructure projects are particularly exposed.
Fix
Treat GIS files as high-value data and apply DLP rules that flag bulk transfers of .shp, .gdb, .kml, .gpx, and .tif files. Restrict GIS server access to named users with logging on every download. For engineering firms: require two-person approval for downloading complete map sets. Western firms holding sensitive infrastructure maps face the same risk from China, Russia, and others.

China-linked spy group has been quietly breaking into government Exchange servers across Asia and one NATO country since 2024

Trend Micro disclosed a China-aligned espionage cluster called SHADOW-EARTH-053 that has been targeting government and defense organizations across South, East, and Southeast Asia plus one NATO European country since at least December 2024. The group breaks in by exploiting unpatched Microsoft Exchange and IIS servers (using known flaws like ProxyLogon), drops a Godzilla web shell for persistent access, then uses DLL sideloading to load ShadowPad - a long-running Chinese implant. The targeting overlaps with Earth Alux and REF7707, suggesting either a shared operator or shared infrastructure across China-aligned groups. Targets include journalists and activists alongside government agencies.

Check
If you run Microsoft Exchange or IIS, confirm every server is patched against ProxyLogon and recent Exchange CVEs - the entry point is unpatched 2-3 year old flaws, not zero-days.
Affected
Government and defense organizations in South, East, and Southeast Asia and the NATO European country are the named targets. Any organization running internet-facing Microsoft Exchange or IIS that has fallen behind on patching is at risk. Diaspora communities and journalists working on China-related stories are at acute risk - the campaign extends transnational repression alongside conventional espionage.
Fix
Patch Microsoft Exchange and IIS to current versions and confirm with active scanning. Hunt for Godzilla web shell artifacts: unusual .aspx files in Exchange's web directories, suspicious POSTs with encrypted payloads, and outbound HTTPS to unfamiliar domains from Exchange/IIS processes. For journalists and activists working on China topics, follow Citizen Lab guidance: hardware MFA, encrypted communications, skepticism of unsolicited story tips.

Two new cybercrime crews are calling employees, getting their MFA codes by phone, then stealing data from SaaS apps within hours

CrowdStrike disclosed two cybercrime groups - Cordial Spider and Snarky Spider - running fast SaaS extortion attacks that stay almost entirely inside legitimate SaaS environments. The pattern: call employees pretending to be IT support, walk them through an 'MFA reset' that's actually a credential-harvesting site that mimics their company's branding, capture the password and MFA code, then immediately log into SSO and pivot through Microsoft 365, Salesforce, and other SaaS apps. The attackers register their own device for MFA and exfiltrate data within hours. Both groups overlap with the broader ShinyHunters ecosystem (UNC6240/UNC6661/UNC6671).

Check
Run a vishing-specific awareness exercise this week. Tell every employee that real IT will never ask them to read out an MFA code over the phone or enter it on a website during a call.
Affected
Organizations with SSO across Microsoft 365, Salesforce, Okta, Google Workspace, or similar SaaS where one set of credentials reaches multiple apps. Acute risk for help-desk-heavy enterprises (financial services, healthcare, large retail) where IT calls feel routine. Any company with a public corporate logo and SSO landing page is in the target pool.
Fix
Make it policy that IT never asks for MFA codes by phone. Require step-up authentication for any MFA registration change. Alert on new MFA device registrations from unfamiliar IPs. In Microsoft 365, monitor for OAuth grants to ToogleBox Recall and similar inbox-rule apps - these were used by Cordial Spider to delete security alerts. Use Mandiant's published IoCs to block known credential-harvesting domains.