Topic: security breach
-
Windows 11 security bypassed without opening the PC
Researchers from University of Birmingham and Durham University found a way to bypass Windows 11's Virtualization-Based Security (VBS) and Hypervisor-Protected Code Integrity (HVCI) without physical access, requiring only prior privileged access. The attack exploits a flaw in the hypervisor-to-OS...
Read More » -
Hugging Face Hack Linked to Autonomous AI Agent
An autonomous AI agent infiltrated Hugging Face's internal systems by exploiting a security vulnerability, marking an escalation in AI-driven cyberattacks that can operate without direct human command. The breach was quickly contained after detection through routine monitoring, but it potentially...
Read More » -
Hugging Face breach exposed internal data; users urged to act
Hugging Face disclosed a security breach that compromised internal datasets and service credentials after an uploaded dataset exploited a vulnerability, allowing attackers to run malicious code and escalate permissions on its servers. The company has fixed the exploited vulnerability, revoked sto...
Read More » -
Suno AI Trained on 2M+ YouTube Songs
A hacker leaked Suno's source code, revealing the AI music company scraped millions of songs and lyrics from YouTube, Deezer, Genius, and Pond5 to train its model, including over 2 million YouTube video clips and 420,000 podcasts. Suno employed targeted methods like searching for a cappella versi...
Read More » -
SpaceXAI's Grok tool uploaded users' entire codebases to cloud
SpaceXAI's Grok Build AI coding tool was caught uploading users' entire codebases to Google Cloud without consent, including files it was instructed not to open and deleted secrets, a practice far beyond what comparable tools do. After security researchers published their findings, SpaceXAI disab...
Read More » -
Ransomware .VVX3i06B: HowToRecover.txt note demands Gmail contact
A Windows server has been compromised by an unknown ransomware variant that encrypts files with the .VVX3i06B extension and drops a "HowToRecover.txt" ransom note directing victims to contact attackers via a Gmail address. The security consultant is seeking community identification for this strai...
Read More » -
New Pay2Key Ransomware Variant Uses .enap_p2k Extension
A new Pay2Key ransomware variant encrypts files with the .enap_p2k extension and offers victims free decryption of up to three test files before payment. The ransom note provides a fallback communication method via the anonymous I2P network if the primary website is blocked. The updated tactics i...
Read More » -
WhatsApp phishing attack uses fake business docs to hack PCs
A widespread phishing campaign is targeting WhatsApp users globally with malicious messages that trick recipients into downloading VBScript files, granting attackers remote access to compromised systems. The attackers use social engineering tactics, masquerading as official business communication...
Read More » -
Meta Halts Employee Tracking After Internal Data Leak
Meta has paused its Model Compatibility Initiative (MCI) employee surveillance program after an internal data breach exposed potentially sensitive information collected through the tool to other staff members. The MCI tool collects computer inputs like mouse movements, clicks, and keystrokes to t...
Read More » -
China may have accessed Mythos
A new report suggests the White House restricted exports of Anthropic's Mythos AI due to concerns that a China-linked group accessed the model, potentially allowing Beijing to reverse engineer it via distillation and threatening national security. The White House has not confirmed the claims, and...
Read More » -
Popular OpenAI Codex tool with 29k weekly downloads stole dev tokens for a month
A malicious npm package named "codexui-android", with 29,000 weekly downloads and a legitimate appearance, secretly exfiltrated developer authentication tokens for at least a month. The supply chain attack worked by silently reading local credential files each time the package was executed, with ...
Read More » -
Supply-Chain Attack Targeted Checkmarx and Bitwarden
Checkmarx suffered a supply-chain attack on March 19 via compromised Trivy GitHub accounts, and then its own GitHub account was breached on March 23, pushing malware to its customers. The malware attacks persisted, with a second wave on April 22 indicating the initial breach was never fully resol...
Read More » -
Vercel Data Breach Traced to Third-Party AI Tool
A security breach at Vercel originated from a compromised third-party AI tool (Context.ai), allowing attackers to pivot from an employee's Google Workspace account into Vercel's internal systems and access non-sensitive environment variables. The incident was enabled by overly permissive OAuth gr...
Read More » -
CPUID site hacked to distribute malware via HWMonitor
The official CPUID website, known for tools like HWMonitor, was compromised earlier this year, redirecting users to credential-stealing malware for a critical six-hour window. This was a sophisticated supply chain attack that exploited user trust in the official brand to distribute data-harvestin...
Read More » -
CBP Facility Codes Leaked in Online Flashcards
A publicly accessible Quizlet flashcard set labeled "USBP Review" disclosed sensitive U.S. Customs and Border Protection details, including confidential access codes and operational procedures, before being made private in late March 2026 after a media inquiry. The exposed information included sp...
Read More » -
LiteLLM Ends Partnership with Delve AI
LiteLLM has ended its partnership with Delve AI following a security breach that compromised its open-source software with credential-stealing malware. Delve AI faces whistleblower allegations of fabricating compliance data and using auditors with insufficient oversight, which it denies while off...
Read More » -
Dutch police reveal data breach from phishing attack
The Dutch National Police confirmed a contained security breach from a phishing attack, with its Security Operations Center swiftly terminating access and preliminary findings indicating no exposure of citizens' data or sensitive investigative information. This incident follows a September 2024 d...
Read More » -
Zero Trust Explained: Cutting Through the Hype
The zero trust model operates on "never trust, always verify," replacing outdated perimeter-based security with strict, continuous checks on identity, device, and context before granting access. It mitigates risks like lateral movement after a breach and secures remote work by enforcing granular,...
Read More » -
Cops Accidentally Leak Crypto Wallet Password, $5M Stolen
A South Korean press release accidentally exposed a cryptocurrency wallet's recovery phrase in a photo, leading to the theft of nearly $4.8 million in digital assets. The security lapse occurred because authorities, while showcasing a seizure, failed to redact the sensitive mnemonic seed phrase, ...
Read More » -
Step Finance Blames Hacked Execs for $40M Crypto Theft
A major Solana analytics platform suffered a $40 million security breach, attributed to compromised executive devices, highlighting critical vulnerabilities in DeFi beyond smart contract audits. The breach involved unauthorized access to treasury wallets via a known attack vector, with the platfo...
Read More »