company logo

Product

Our Product

We are Reshaping the way Developers find and fix vulnerabilities before they get exploited.

Solutions

By Industry

BFSI

Healthcare

Education

IT & Telecom

Government

By Role

CISO

Application Security Engineer

DevsecOps Engineer

IT Manager

Resources

Resource Library

Get actionable insight straight from our threat Intel lab to keep you informed about the ever-changing Threat landscape.

Subscribe to Our Weekly Threat Digest

Company

Contact Us

Have queries, feedback or prospects? Get in touch and we shall be with you shortly.

loading..
loading..
loading..
Loading...

iCloud

loading..
loading..
loading..

iCloud Calendar Abused to Send Phishing Emails from Apple’s Servers

Attackers exploit iCloud Calendar invites via Apple servers to deliver phishing scams.

13-Sep-2025
7 min read

No content available.

Related Articles

loading..

RCE

Actively exploited CVE-2025-21043 lets attackers run code via Samsung’s Quram im...

Samsung fixed a **critical remote-code-execution** bug in the Quramsoft image codec (`libimagecodec.quram.so`) used on Galaxy devices. The flaw—an **Out-of-Bounds Write**—is patched in **SMR Sep-2025 Release 1** and was **exploited in the wild** before disclosure. Samsung credits **Meta & WhatsApp Security Teams** (report date **2025-08-13**). **Patch immediately** to the **September 2025** security level (or later). Media coverage underscores that **any messenger or app that relies on the system image parser** (not just WhatsApp) could be a delivery vector for a malicious image. ### Vulnerability details * **CVE:** CVE-2025-21043 * **Component:** Quramsoft image codec (`libimagecodec.quram.so`) * **Type / CWE:** Out-of-Bounds Write (**CWE-787**) → memory corruption → potential RCE * **Severity:** **Critical** (Samsung) * **Affected OS on Samsung builds:** **Android 13–16** * **Fix:** Included in **SMR Sep-2025 Release 1** * **Exploit status:** **“Exploit … has existed in the wild.”** (Samsung) * **Reporter credit:** **Meta & WhatsApp Security Teams** (reported **2025-08-13**) All of the above are specified in Samsung’s security bulletin entry for SVE-2025-1702 / CVE-2025-21043. ([Samsung Mobile Security][1]) > **Why it matters:** Image codecs often process **untrusted content automatically** (previews, thumbnailers, notifications). One malformed image can crash the decoder or, with a working exploit, **run attacker code** in the decoder’s process. Coverage notes **other messengers using the system codec** may be exposed—not just WhatsApp. ([BleepingComputer][2]) ### Timeline (UTC) * **2025-08-13:** Samsung receives report from **Meta & WhatsApp**. ([Samsung Mobile Security][1]) * **2025-09-03 → 2025-09-12:** **September 2025 SMR** announced and begins rolling out to models/regions (ongoing staggered release). Community trackers and press confirm early-September waves across multiple devices. ([SamMobile][3]) * **2025-09-12:** Public reporting highlights active exploitation and WhatsApp/Meta credit. ([BleepingComputer][2]) ### Threat model & likely attack paths **Primary vector:** Delivery of a **malicious image** that triggers decoding paths eventually invoking `libimagecodec.quram.so`. Apps that **delegate image parsing to the OS/vendor codecs** inherit the risk. **Common touchpoints** (high-exposure first): 1. **Messaging**: MMS/RCS; OTT messengers (auto-download, previews, notification thumbs). 2. **System components**: Gallery, MediaScanner, indexers/thumbnailers, ShareSheet. 3. **Browsers & Email**: Inline image rendering, preview panes. BleepingComputer emphasizes that **other messengers besides WhatsApp** may be susceptible if they use the system codec. ([BleepingComputer][2]) --- # 4) Root-cause clarity (what’s known vs. inferred) * **Confirmed (Samsung):** **Out-of-Bounds Write** in the Quramsoft codec; fixed by correcting an “incorrect implementation.” No format or trigger details disclosed. ([Samsung Mobile Security][1]) * **Context (inference, not specific to this CVE):** Historically, Quram/Qmage bugs involved **size/stride miscalculation** and **insufficient metadata validation**, enabling heap corruption and even **zero-click MMS** attacks on Samsung devices. Treat this as background, not proof for CVE-2025-21043. ([Project Zero][4]) **Exploitability factors to keep in mind (general):** Modern Android hardening (ASLR, CFI, PAC/MTE on some SoCs, process sandboxing) raises difficulty but **doesn’t preclude** reliable RCE chains. Exact exploit quality/sandbox impact here are **undisclosed**. --- # 5) Impact assessment * **Worst-case:** **Network-reachable RCE** when the image is parsed, leading to code execution within the decoder’s process context. Additional bugs or misconfigurations would be required for broader device compromise. (RCE status per Samsung.) ([Samsung Mobile Security][1]) * **Breadth:** Any Galaxy device on **pre-SMR Sep-2025** builds and using vendor image paths is at risk until patched; rollout is staggered by model/region/carrier. Trackers highlight that **September SMR** is actively shipping. ([SamMobile][3]) --- # 6) Exposure verification & validation ## 6.1 “Am I safe?” quick check (on-device) * Go to **Settings → About phone → Android version → Android security update**. * If it shows **September 1, 2025** or **September 5, 2025** (or a later month), you have the corresponding Android patch level; Samsung’s SMR includes these plus Samsung fixes. Google states that **2025-09-05** includes all issues in **2025-09-01** and earlier. ([Android Open Source Project][5]) * Samsung’s bulletin explicitly says **SMR Sep-2025 Release 1** includes the fix for **CVE-2025-21043**. ([Samsung Mobile Security][1]) > **Rule of thumb:** On Samsung, a device showing **“September 2025”** with **SMR Sep-2025 R1** is **covered** for CVE-2025-21043. ## 6.2 ADB/MDM checks (fleet) * **Read patch string:** `adb shell getprop ro.build.version.security_patch` → expect `2025-09-01` or `2025-09-05` (or later). ([Android Open Source Project][5]) * **Model-specific rollout:** Confirm availability for particular SKUs/regions using Samsung community/tracker posts (e.g., **S23 series S918BXXS8DYI3** reported live). Use as **availability signal**, not authoritative coverage. ([Droid Life][6]) --- # 7) Detection & triage guidance (blue team) > Focus on **decoder crashes** and **media-ingestion correlations**. Don’t expect IOCs yet; content is often **E2EE**. 1. **Crash/tombstone review on test or enrolled devices** * Pull `/data/tombstones/` (`tombstone_*.txt`). * Look for processes loading **`libimagecodec.quram.so`** with crashes around image receipt/preview (e.g., **SIGSEGV/SIGABRT**, heap corruption markers). * Correlate timestamps with incoming media (messaging push, notification arrivals, gallery indexing). 2. **Logcat spot-checks** (developer/test devices) * `adb logcat | grep -iE "libimagecodec\.quram\.so|Fatal signal|backtrace"` 3. **MDM/EDR telemetry (where present)** * Alerts on repeated **media decode faults**, **abnormal child processes** spawned by media frameworks, or **RWX/JIT anomalies** inside media processes. 4. **Threat intel tracking** * Monitor the **Samsung SMR page** for updates; watch reporting threads for sample hashes or exploit indicators once (if) they become public. ([Samsung Mobile Security][1]) --- # 8) Immediate mitigations & hardening **Top priority: patch to SMR Sep-2025 R1+ now.** Users: **Settings → Software update → Download and install**. Enterprises: push via **MDM** compliance policy. ([Samsung Mobile Security][1]) **Until patched:** * **Messaging controls:** Disable **auto-download/auto-preview** of images where policy allows; restrict unknown senders. (Media outlets warn other messengers using system codecs could be targeted.) ([BleepingComputer][2]) * **MDM baselines:** Require **security patch ≥ 2025-09-01** (prefer **2025-09-05**), block devices below that level; enforce **Play Protect**, restrict sideloading. ([Android Open Source Project][5]) * **Network caveat:** For E2EE messengers, server-side content rewrites/stripping won’t apply; rely on **endpoint** controls and rapid patching. --- # 9) Developer guidance (apps & enterprise toolmakers) * If your app **defers to system image decoders** (Java/Kotlin `BitmapFactory`, `ImageDecoder`, or native via NDK that calls into vendor libs), your app inherits platform risk. Ship **app updates** promptly (even if only to gate risky auto-previews behind user action) until fleet patching reaches critical mass. Coverage flags that **system-codec users** are exposed. ([BleepingComputer][2]) * Avoid introducing alternate native image parsers **unless** they are proven, memory-safe, and well-maintained—rolling your own parser typically **increases** risk. * Expand **fuzzing** against image inputs in CI and consider **content triage** (e.g., size/dimension caps) before decoding. Historical Quram/Qmage work shows how metadata misparsing leads to corruption. ([Project Zero][4]) --- # 10) Validation plan (post-patch) 1. **Confirm patch level** (Section 6). 2. **Negative testing:** Re-ingest previously crashing images (if any were captured internally) on a **fully patched** test device. **Do not** share or re-distribute suspect samples. 3. **Stability watch:** No new tombstones for media pipelines over 72h of normal use across patched pilot group. 4. **Close incident** when fleet compliance ≥ agreed threshold (e.g., **95% patched**), with exception-handling for stragglers. --- # 11) Risk scoring & policy * **Severity:** **Critical** (vendor). Treat as **Priority-1** because of **confirmed in-the-wild exploitation**. Samsung didn’t publish a CVSS; don’t rely on third-party estimates for gating decisions. ([Samsung Mobile Security][1]) --- # 12) Open questions (track & update) * Which **image formats** and specific parser paths are affected? * **Triggering requirements** (zero-click via auto-thumbnailing vs. open)? * **Exploit chain** details (sandbox escape/priv-esc)? * **Public technical write-ups / PoC?** (None official at time of writing.) Samsung’s bulletin and reputable reporting are the **authoritative** sources for now. ([Samsung Mobile Security][1]) --- # 13) Ops playbooks & copy-paste blocks ## 13.1 End-user comms (short) > We’re deploying the **September 2025 security update** to Galaxy devices due to a **critical image-parsing vulnerability (CVE-2025-21043)** that is **actively exploited**. Please go to **Settings → Software update → Download and install** and apply the update today. Until you’re patched, avoid opening images from unknown contacts and disable auto-download of media in messaging apps. ([Samsung Mobile Security][1]) ## 13.2 SOC hunt checklist * [ ] Pull recent **tombstones**; flag crashes mentioning `libimagecodec.quram.so`. * [ ] Correlate with **inbound image receipt** times in messengers. * [ ] Triage any **repeatable** crash sequences on **pre-patch** devices; escalate if persistence or unusual child processes observed. * [ ] Track fleet **patch compliance** daily until ≥ 95%. ## 13.3 MDM compliance policy * **Minimum Android security patch:** **`2025-09-05`** (preferred) or **`2025-09-01`** (temporary). Devices below are **non-compliant**. (Patch-level semantics per Google’s ASB.) ([Android Open Source Project][5]) * **Controls:** Disable image auto-download in corporate messengers (where configurable), enforce **Play Protect**, block sideloading, and require **full-disk encryption**. ## 13.4 Engineer notes (test devices) * **Read patch level:** `adb shell getprop ro.build.version.security_patch` → expect `2025-09-0X`. ([Android Open Source Project][5]) * **Crash triage:** `adb logcat | grep -iE "libimagecodec\.quram\.so|Fatal signal|backtrace"` * **(Optional)** Inspect presence/paths of vendor codec libs: `adb shell ls -l /vendor/lib64/libimagecodec.quram.so || ls -l /system/lib64/libimagecodec.quram.so` (Presence alone ≠ vulnerable/patched; use **patch level** as the source of truth—Samsung doesn’t publish per-file versions in the bulletin.) ([Samsung Mobile Security][1]) --- # 14) “Am I patched?” quick table | Device shows… | What it means | Action | | --------------------------------------- | ------------------------------------------- | --------------------------------------------------------------------------------------------- | | **Android security update: 2025-09-05** | You have the complete September patch level | ✅ Good; verify SMR says September 2025 | | **Android security update: 2025-09-01** | You have the baseline September patch level | ✅ Temporarily acceptable; still includes SMR Sep-2025 R1 on Samsung if offered for your model | | **August 2025 or earlier** | Vulnerable until OEM SMR is installed | ❌ Update immediately | Google defines the **01/05** patch levels; Samsung’s **SMR Sep-2025 R1** is the vendor package that contains the **CVE-2025-21043** fix. ([Android Open Source Project][5]) --- # References (primary) * **Samsung Mobile Security — SMR Sep-2025 Release 1** (lists **CVE-2025-21043**, **Critical**, **Android 13–16**, **reported 2025-08-13**, and **“exploit … existed in the wild.”** Also credits **Meta & WhatsApp Security Teams**). ([Samsung Mobile Security][1]) * **BleepingComputer** — reports active exploitation, WhatsApp/Meta credit, and risk to other messengers using the system codec. ([BleepingComputer][2]) * **Android Security Bulletin — September 2025** — explains **2025-09-01/05** patch-level semantics for validation. ([Android Open Source Project][5]) * **Rollout context** — Samsung September SMR announcement/tracker posts showing early-September waves to devices (e.g., S23). ([SamMobile][3]) * **Historical background** — Project Zero’s **Qmage** research on Samsung image codecs and attack surface (context only). ([Project Zero][4]) --- ## Want device-specific guidance? Tell me your **exact Galaxy model** and the **security patch level** it shows in **Settings**; I’ll map it to current rollout notes and tell you whether you’re covered or need to escalate patching. [1]: https://security.samsungmobile.com/securityUpdate.smsb "Samsung Mobile Security" [2]: https://www.bleepingcomputer.com/news/security/samsung-patches-actively-exploited-zero-day-reported-by-whatsapp/ "Samsung patches actively exploited zero-day reported by WhatsApp" [3]: https://www.sammobile.com/news/samsung-monthly-updates-september-2025-security-patch-is-a-big-one/?utm_source=chatgpt.com "Samsung monthly updates: September 2025 security patch ..." [4]: https://googleprojectzero.blogspot.com/2020/07/mms-exploit-part-1-introduction-to-qmage.html?utm_source=chatgpt.com "MMS Exploit Part 1: Introduction to the Samsung Qmage ..." [5]: https://source.android.com/docs/security/bulletin/2025-09-01?utm_source=chatgpt.com "Android Security Bulletin—September 2025" [6]: https://www.droid-life.com/2025/09/11/samsung-september-updates-roll-out/?utm_source=chatgpt.com "Samsung September Updates Roll Out to These Devices"

loading..   12-Sep-2025
loading..   10 min read
loading..

DDoS

FastNetMon detects record 1.5 Gpps UDP flood from 11,000+ compromised CPEs, expo...

FastNetMon has confirmed detection of a **record-scale Distributed Denial of Service (DDoS) attack**, peaking at **1.5 billion packets per second (Gpps)**. The assault, targeting a European DDoS scrubbing provider, is one of the highest packet-rate floods ever disclosed publicly. While not the largest in raw bandwidth, the event highlights the evolving threat of **packet-saturation attacks** designed to overwhelm router CPU, control planes, and scrubbing pipelines rather than transit links. ## Attack Profile The attack was not a conventional volumetric flood but a **massive UDP-based packet storm** launched from globally distributed compromised devices. * **Vector**: UDP flood targeting scrubbing infrastructure. * **Scale**: 1.5 Gpps peak rate. * **Bandwidth**: Lower than multi-Tbps volumetric floods, but optimized for packet-per-second impact. * **Origin**: Over **11,000 autonomous networks** contributed traffic. * **Botnet composition**: Large clusters of compromised **MikroTik routers and IoT devices**, acting as customer premises equipment (CPE). This methodology indicates adversaries are prioritizing **state-exhaustion vectors** over pipe saturation, aiming at **router forwarding engines, ACL tables, and scrubbing CPUs**. ## Detection Mechanisms [FastNetMon](https://fastnetmon.com/2025/09/09/press-release-fastnetmon-detects-a-record-scale-ddos-attack/)’s **flow telemetry analysis platform**, written in optimized C++, enabled near-real-time detection. Critical elements: * **Flow-based anomaly detection**: Identifying packet-rate anomalies at the millisecond scale. * **CPU-efficient algorithms**: Parsing billions of NetFlow/IPFIX records without loss. * **Real-time signaling**: Immediate trigger of defensive ACLs and scrubbing workflows. Detection latency is crucial in Gpps-scale attacks, where control plane resources can be exhausted within seconds. ## Mitigation Strategies Mitigation combined automated filtering, rate-limiting, and scrubbing workflows: * **Access Control Lists (ACLs)** deployed on upstream edge routers to discard obvious spoofed traffic. * **Scrubbing center packet inspection**, including UDP state analysis and anomaly detection. * **Dynamic blackholing** for unmitigable subnets, sacrificing reachability to preserve upstream stability. * **Rate-limit enforcement** for specific UDP ports leveraged in amplification or flood scenarios. While effective in this instance, the reliance on ACLs at such scale exposes **edge router performance bottlenecks**. ACL deployment at 1.5 Gpps stresses TCAM capacity and control plane update cycles. ## Strategic Implications This event represents a **paradigm shift** in attacker priorities: * **From bandwidth to packet-rate**: Threat actors are engineering floods to stress *packet processing pipelines*, bypassing traditional bandwidth-centric defenses. * **CPE exploitation**: The reliance on compromised routers and IoT devices underscores persistent **firmware negligence** and **default credential exploitation**. * **ISP responsibility**: Filtering at the **ISP level** is essential; without outbound UDP controls, infected CPEs become unfiltered launchpads. * **Scrubbing resilience**: Providers must scale to handle not just Tbps floods but **multi-Gpps packet rates**. ## Comparative Context Other recent incidents highlight the dual evolution of attack methodologies: * **[Cloudflare](https://www.secureblink.com/cyber-security-news/cloudflare-crushes-11-5-tbps-d-do-s-blitz-from-google-cloud) mitigation (2025)**: Reported **11.5 Tbps, 5.1 Bpps** floods, prioritizing volumetric scale alongside packet-rate stress. * **[FastNetMon](https://fastnetmon.com/2025/09/09/1-5-billion-packets-per-second-ddos-attack-detected-with-fastnetmon/) detection (2025)**: Emphasis on packet-rate intensity at 1.5 Gpps, demonstrating adversaries can weaponize **smaller devices at massive distribution scale**. **Comparison Snapshot**: | Attack Event | Scale | Nature | Key Target | Implication | | ----------------- | -------------------- | -------------------- | ---------- | ------------------------ | | Cloudflare (2025) | 11.5 Tbps / 5.1 Bpps | Volumetric + packets | Edge pipes | Bandwidth exhaustion | | FastNetMon (2025) | 1.5 Gpps | Pure packet flood | Scrubbing | Control plane exhaustion | This dual trend suggests defenders must build **multi-layered resilience**: bandwidth mitigation *and* packet-rate scaling. ## Recommendations ### For ISPs * Implement **egress filtering** (BCP38/BCP84) to suppress spoofed UDP from customer networks. * Deploy **telemetry pipelines** for per-subscriber packet-rate anomaly detection. * Maintain **ACL automation frameworks** capable of near-instant deployment at line rate. ### For Scrubbing Providers * Architect scrubbing centers with **packet-rate scaling in mind**, not just raw bandwidth capacity. * Offload filtering to **programmable ASICs and FPGA-based platforms** to avoid CPU bottlenecks. * Invest in **low-latency telemetry triggers** that initiate mitigation before exhaustion thresholds are hit. ### For Device Vendors & Enterprises * Enforce **secure defaults**: no factory default passwords, minimal UDP exposure. * Ensure **firmware patch pipelines** for consumer routers and IoT devices. * Promote **carrier-grade automatic update mechanisms** for widely deployed CPE. The 1.5 Gpps flood detected by FastNetMon is a **milestone in DDoS evolution**. It highlights a new generation of threats where **packet processing exhaustion** is prioritized over bandwidth saturation. With IoT and CPE devices weaponized into global botnets, the defensive burden shifts to ISPs, scrubbing providers, and device vendors alike. Without **systemic adoption of ISP-level egress filtering, firmware hardening, and packet-rate aware scrubbing infrastructure**, the next wave of Gbps-scale floods could cripple even the most prepared networks. This incident should be treated not as an outlier but as a **preview of the coming normal** in high-velocity, distributed denial-of-service warfare.

loading..   11-Sep-2025
loading..   4 min read
loading..

NPM

Supply Chain

Hijacked npm packages with 2.6B weekly downloads spread crypto-stealing malware,...

JavaScript ecosystem experienced one of its most severe supply-chain compromises to date. Attackers infiltrated the npm account of a prominent maintainer and inserted malicious code into **18 popular packages**, including *chalk*, *debug*, and *ansi-styles*. Together, these libraries are downloaded more than **2.6 billion times every week**, meaning the poisoned code had the potential to propagate across countless applications, frameworks, and enterprises worldwide. Unlike previous incidents where malware reached [npm](https://www.secureblink.com/cyber-security-news/top-npm-package-is-hijacked-secret-malware-spreads-in-dev-supply-chain-attack) through newly published typosquats, this breach leveraged the credibility of **high-trust maintainers**. The attackers’ strategy demonstrates both the fragility of open-source distribution channels and the sophistication of modern adversaries targeting developers. ## How the Attack Transpired The breach began with a **spear-phishing email** sent to [Josh Junon](https://bsky.app/profile/bad-at-computer.bsky.social/post/3lydmyzpwa22s) (known online as *[Qix](https://www.npmjs.com/~qix)*), maintainer of several foundational JavaScript packages. The email came from a spoofed domain, `npmjs.help`, and [warned](https://bsky.app/profile/bad-at-computer.bsky.social/post/3lydioq5swk2y) of an impending account lockout scheduled for September 10. To “resolve” the issue, the maintainer was directed to a convincing clone of the official [npm](https://bsky.app/profile/bad-at-computer.bsky.social/post/3lydje4zqis2y) login portal. Once credentials and two-factor authentication codes were harvested, the attackers logged into the legitimate npm account and began publishing new versions of widely used packages. Because the versions appeared authentic and came from a trusted maintainer, automated dependency resolution tools quickly propagated the tainted updates into developer environments. ## Malicious Payload Analysis of the injected code reveals a **lightweight but effective cryptocurrency stealer**. Hidden in initialization routines, the payload executed only under specific conditions: * **Targeted Execution**: Triggered when `NODE_ENV=production`, reducing the chance of detection during development. * **Network Hooking**: Intercepted `XMLHttpRequest` and `WebSocket` calls to wallet providers such as MetaMask. * **Transaction Hijacking**: Substituted legitimate destination wallet addresses with attacker-controlled ones. * **Stealth Features**: No external requests or large dependencies were added, minimizing observable changes in behavior. This design shows an acute awareness of how developers audit code: subtle, conditional, and deeply embedded, it blended seamlessly with legitimate logic. ## Scope of Impact The scale of potential impact is staggering. *chalk* alone is a ubiquitous dependency for styling terminal [output](https://github.com/orgs/community/discussions/172738), embedded in countless tools from build systems to testing frameworks. *debug* underpins logging across major Node.js applications. Through **transitive dependencies**, millions of developers and organizations could have unknowingly introduced the malware into their build pipelines and runtime environments. While npm security teams acted quickly to unpublish the compromised versions, the window of exposure—measured in hours, not weeks—was sufficient for global uptake. Whether the payload successfully redirected cryptocurrency transactions at scale remains under investigation. This attack is not just another data point in the long history of supply-chain compromises. It underscores several systemic issues: 1. **Human Factors Outpace Technical Safeguards** Even with two-factor authentication, phishing can capture live codes and bypass protections. Hardware keys with WebAuthn provide far stronger resistance. 2. **Dependency Chains Multiply Risk** Developers rarely import these packages directly. Instead, they flow transitively through larger frameworks. That makes risk assessment and mitigation far harder. 3. **Trust Model Limitations** npm’s security relies heavily on the assumption that verified maintainers will remain uncompromised. This single point of trust creates a wide attack surface. ## Defensive Measures For developers and organizations concerned about exposure, immediate steps include: * **Audit and Pin**: Check dependency trees for the specific compromised versions and downgrade or pin to safe releases. Avoid using loose semantic ranges until ecosystems stabilize. * **Use Internal Registries**: Proxy packages through controlled artifact repositories like Artifactory, Nexus, or Verdaccio. This ensures malicious versions can be quarantined before internal use. * **Adopt Runtime Scanning**: Apply sandbox analysis or dynamic behavior detection to new dependencies before they reach production environments. * **Strengthen Authentication**: Encourage maintainers and internal teams to migrate from TOTP-based 2FA to phishing-resistant hardware keys. At the ecosystem level, greater adoption of **cryptographic package signing (e.g., Sigstore)** could provide an additional verification layer, ensuring consumers can detect tampering even if a maintainer account is compromised.

loading..   09-Sep-2025
loading..   4 min read