All news with #information disclosure tag

🚨 Balancer announced an exploit of its V2 Compostable Stable Pools on Ethereum at 07:48 UTC that resulted in reported losses exceeding $128 million. Initial analysis from GoPlus Security points to a precision rounding error in the Vault’s swap calculations that an attacker chained via batchSwap, while other researchers suggest improper authorization and callback handling in V2 vaults. Balancer says the issue is isolated to V2 Compostable Stable Pools, with V3 and other pools unaffected, and the team is working with security researchers on a full post‑mortem. Users are warned to remain vigilant for scams and phishing attempts following the incident.

🔒 Microsoft has issued updates to address three previously unknown flaws in the Windows Graphics Device Interface (GDI) that could permit remote code execution and information disclosure. The issues, rooted in malformed EMF/EMF+ records, cause out-of-bounds memory access in GdiPlus.dll and gdi32full.dll during image rendering, thumbnailing and print initialization. Patches were released across the May, July and August 2025 Patch Tuesdays (KB5058411, KB5062553, KB5063878); administrators should apply updates promptly and avoid opening untrusted EMF files.

⚠️ CISA disclosed critical vulnerabilities in Vertikal Systems Hospital Manager Backend Services that were fixed as of September 19, 2025. One flaw exposed the unauthenticated ASP.NET tracing endpoint (/trace.axd), allowing disclosure of request traces, headers, session identifiers, and internal paths. A second flaw returned verbose ASP.NET error pages for invalid WebResource.axd requests, revealing framework versions, stack traces, and server paths. CVE-2025-54459 and CVE-2025-61959 were assigned; organizations should apply vendor updates and follow network isolation best practices.

🔎 Google Threat Intelligence Group (GTIG) observed coordinated pro-Russia information operations responding to reported Russian drone incursions into Polish airspace on Sept. 9–10, 2025. Actors amplified narratives denying Russian culpability, blaming NATO or Poland, and seeking to erode domestic and international support for Ukraine. GTIG documented activity across multiple networks and languages and noted these operations leveraged both long-standing and recently developed influence infrastructure.

⚠️ Researchers have disclosed Pixnapping, a pixel-stealing side-channel that can extract 2FA codes, Maps timelines, and other sensitive UI contents from Android apps by abusing GPU compression together with Android's window-blur and intent mechanisms. The proof-of-concept captures codes in under 30 seconds on several Google and Samsung devices running Android 13–16 without requiring special manifest permissions. Google tracked the issue as CVE-2025-48561 (CVSS 5.5) and issued mitigations in the September 2025 Android Security Bulletin, but researchers say a workaround can re-enable the technique and that some app-list bypass behavior will not be fixed.

⚠️ Oracle released an emergency security alert on October 11 for CVE-2025-61884, a 7.5 CVSS information-disclosure flaw in the Runtime UI component of E-Business Suite (versions 12.2.3–12.2.14). The vulnerability allows unauthenticated remote attackers with network access to steal sensitive data. The patch arrives one week after an emergency fix for a Cl0p-exploited RCE, and experts urge administrators to apply updates, hunt for prior compromise, and restrict outbound traffic from EBS servers.

🔒 Oracle released an emergency update to address CVE-2025-61884, an information disclosure flaw in the E-Business Suite Runtime UI that affects versions 12.2.3 through 12.2.14. The vulnerability is remotely exploitable without authentication and has been assigned a CVSS base score of 7.5, meaning a successful exploit could expose sensitive resources. Oracle strongly urges customers to apply the out-of-band patch or recommended mitigations immediately, particularly for internet-facing instances.

🖱️ Researchers at the University of California, Irvine demonstrated a proof-of-concept called Mic-E-Mouse, showing that high-end optical mice can pick up desk-transmitted voice vibrations and be used to reconstruct nearby conversations. The attack can be executed on PC, Mac and Linux by non-privileged user-space programs, and Wiener and neural-network filtering was used to enhance muffled signals into intelligible speech. Practical limits include a quiet environment, thin desks (≈3 cm or less), mostly stationary mice and very high-DPI hardware; placing a rubber pad or mouse mat under the mouse prevents the leakage.

🔒 Broadcom released updates addressing two high-severity VMware NSX vulnerabilities reported by the U.S. National Security Agency (NSA). The flaws (CVE-2025-41251 and CVE-2025-41252) permit unauthenticated attackers to enumerate valid usernames via a weak password-recovery flow and a separate enumeration vector, which could be used to support brute-force or unauthorized login attempts. Administrators should apply the vendor patches immediately and verify recovery workflows and logging.

⚠️ Festo devices running affected EtherNet/IP firmware are vulnerable to multiple remotely exploitable issues, including incorrect numeric conversions, out-of-bounds reads, and reachable assertions that can lead to denial-of-service or data disclosure. Combined CVSS scores reach up to 8.2, and successful exploitation requires low attack complexity. Festo reports no planned fixes; CISA advises minimizing network exposure, disabling EtherNet/IP when unused, isolating control networks, and using secure remote access such as up-to-date VPNs. Organizations should limit exposure, monitor EtherNet/IP activity, and report suspected incidents.

🔓 Researchers at ETH Zurich disclosed VMScape, a Spectre-like speculative-execution attack that lets a malicious VM extract secrets from an unmodified QEMU hypervisor running on many modern AMD and some Intel CPUs. The exploit abuses shared branch-prediction structures and a FLUSH+RELOAD side channel to induce speculative disclosure. It works without host compromise and bypasses default mitigations; vendors and Linux developers released advisories and kernel patches to mitigate the issue.

🛡️ CISA republished information from Siemens ProductCERT regarding two vulnerabilities affecting the RUGGEDCOM RST2428P (6GK6242-6PA00). The issues — uncontrolled resource consumption (CVE-2025-40802) and exposure of sensitive information (CVE-2025-40803) — are exploitable from an adjacent network and have low CVSS scores (v3.1=3.1; v4=2.3). Siemens recommends firewalling UDP discovery ports and following industrial security guidance; CISA advises minimizing network exposure and isolating control networks.

🔍 The UpGuard Cyber Risk Team discovered a publicly readable Amazon S3 bucket containing spreadsheets that appeared to describe GoDaddy infrastructure running in the AWS cloud. The largest file listed more than 24,000 hostnames and 41 configuration fields, including hostname, OS, workload, region, vCPU, memory and modeled cost data, plus apparent AWS discount information. While the files did not contain credentials or end-user data, they effectively mapped a large-scale cloud deployment and revealed sensitive pricing details. UpGuard notified GoDaddy, and the exposure was closed after coordination with the company.

🔓 The UpGuard Cyber Risk Team discovered a publicly accessible AWS S3 bucket containing a 1.2 TB ndjson file with 48 million records belonging to LocalBlox. The dataset included names, addresses, dates of birth, scraped LinkedIn and Facebook content, Twitter handles, and blended data from sources like Zillow. UpGuard notified LocalBlox on February 28, 2018, and the bucket was secured the same day. This exposure highlights the real-world risk of simple cloud misconfigurations.

🔓 UpGuard researchers discovered an Amazon S3 bucket tied to the Democratic Senatorial Campaign Committee exposing a 145MB zip file that contained a CSV of roughly 6.2 million email addresses. The unprotected bucket granted global authenticated FULL_CONTROL, allowing anyone with a free AWS account to access or modify contents. The file, last modified in 2010 and named EmailExcludeClinton.csv, appears to be an exclusion list and includes consumer, .edu, .gov, and .mil domains. UpGuard notified DSCC and the bucket was secured the following day.

🔒 UpGuard discovered an unauthenticated Elasticsearch instance exposing roughly 22 million web-request records tied predominantly to Leakzone, a forum for illicit data and hacking tools. The logs contained domains, client IPs, geolocation and ISP metadata, and request sizes spanning late June through the July 2025 discovery. Analysis shows widespread use of public proxies and VPN exit nodes, with much traffic routed through major cloud providers, limiting reliable geolocation.