All news with #browser security tag

🔍 Microsoft discovered a malicious Chrome extension posing as Perplexity that logged every search query and each character typed in the address bar by routing input through an attacker-controlled server before redirecting to legitimate results. The extension, named "Search for perplexity ai" and using a look-alike domain, set itself as the default search engine and redirected queries and live suggestion traffic to the attacker domain, collecting headers, IPs, and user agent data. Microsoft reported the extension to Google, which removed it from the Chrome Web Store; defenders are urged to remove the extension and verify search settings immediately.

🔍 Microsoft Threat Intelligence discovered a malicious Chromium extension impersonating Perplexity AI to intercept Omnibox queries and real-time search suggestions. The extension used MV3, declarativeNetRequest rules, and a typosquatted domain (perplexity-ai[.]online) to route searches through attacker infrastructure before redirecting to expected providers. Google removed the extension after responsible disclosure. Microsoft provides indicators, dynamic analysis findings, and mitigation guidance.

🛡️ A malicious Microsoft Edge extension named Edgecution was used to bypass the browser sandbox and deploy a Python-based backdoor by abusing the Chrome Native Messaging protocol. Attackers lured victims via fake Microsoft update pages and social engineering on Microsoft Teams, delivering a malformed ZIP with an embedded Python runtime and two components: a headless Edge extension and a native Python backdoor. Zscaler links the activity to an IAB associated with the Payouts Kings ransomware operation and provides IoCs and mitigation recommendations.

🛡️ Researchers at Graz University of Technology describe FROST, a novel side-channel technique that uses the browser's origin private file system (OPFS) to monitor SSD timing and infer user activity. A malicious webpage leveraging OPFS can repeatedly access storage, measuring micro-delays that reveal what apps or websites are active. The team demonstrated data transmission rates around 660–720 bits/s with ~90% accuracy and used AI to classify app and site fingerprints. Practical constraints — RAM caching, large file creation, and likely EDR/XDR detection — limit FROST to targeted attacks, but it highlights hardware-level blind spots in modern security.

🔒 Menlo Security's 2026 Browser Threat Report warns that many cybersecurity products fail to detect browser-based attacks. Based on telemetry from millions of enterprise browser sessions between January and March 2026, the research found one in five phishing attacks targeting enterprise browser users went undetected by legacy tools. The report highlights that modern enterprise activity increasingly occurs inside browsers, creating blind spots for products not built for the browser session layer. Menlo urges organizations to govern the browser session layer to better protect users and AI agents.

🛡️ Researchers at Graz University of Technology describe FROST, a browser-based timing attack that uses the Origin Private File System (OPFS) to infer which sites a user visits and which apps they open. The exploit runs purely in JavaScript, requires no native code or permissions, and sharpens timer resolution via cross-origin isolation. On macOS it achieved high fingerprinting accuracy, while mitigations remain limited and browser vendors have not implemented firm fixes.

🔒 Brave Software released Brave Origin, a paid, minimalist edition of its browser that omits cryptocurrency, AI, rewards, and monetization-focused features. The company positions Origin for users seeking a streamlined, privacy-focused experience while retaining core protections like Brave Shields. Origin is available as a standalone download or as an upgrade for existing installations, priced at a one-time $59.99 for up to 10 devices (free on Linux).

🛡️ Chromium contains an unpatched vulnerability that lets attackers keep a Service Worker alive across restarts and execute JavaScript persistently. Reported by researcher Lyra Rebane, the bug abuses the Background Fetch API and a race that creates and aborts background fetches to evade UI visibility. Although some UI fixes were applied in 2023, the deeper issue—preventing indefinite Service Worker lifetimes—remains unresolved and can enable tracking, crypto mining, and browser-based bots.

🔒 Security researchers warn of a new scareware strain, CypherLoc, used in around 2.8 million attacks since early 2026. The campaign starts with phishing that directs victims to a malicious page which only activates when specific URL fragments and cryptographic checks pass. Once triggered, the code forces full-screen browser lockdowns, disables controls, displays fake security warnings and a fraudulent support number, with operators posing as Microsoft support. Barracuda urges anti-phishing, browser and endpoint protections and user education to mitigate the threat.

🔒 Bridewell's Cyber Threat Intelligence Report 2026 warns that attackers are abandoning traditional malware for browser- and identity-focused techniques such as ClickFix, FileFix and ConsentFix that trick users into approving commands or authentication prompts. These tactics bypass endpoint controls and MFA because they operate within trusted workflows and are harder to detect. The firm urges stronger identity protection, user awareness and threat-informed defence.

⚠️ Researchers at LayerX Security disclosed a flaw dubbed ClaudeBleed in Anthropic’s Claude in Chrome extension that lets other extensions inject scripts and commandeer the assistant. The issue stems from an exposed messaging interface that trusts origins instead of execution context, enabling zero-permission extensions to issue prompts and perform cross-site actions. Anthropic released a partial patch (v1.0.70) on May 6; LayerX urges stronger mitigations.

🔒 Traditional DLP tools often miss significant risk that originates inside the web browser. The piece explains how users routinely copy, type, and upload sensitive data into web forms, AI prompts, and unsanctioned accounts—activities that endpoint, network, and cloud DLP commonly cannot inspect with the needed context. Browser-native DLP operates within the browsing session to detect clipboard events, form inputs, and uploads, understand which app and account are involved, and apply inline controls to block or warn on risky actions while preserving approved workflows.

🔓 Researchers found that a new infostealer, VoidStealer, can bypass Chrome’s App-Bound Encryption by attaching to the browser process as a debugger and setting breakpoints at decryption routines. At the moment the browser decrypts data, the malware reads the master key directly from memory, enabling theft of session cookies and other secrets. The technique affects other Chromium-based browsers and is available as malware-as-a-service, increasing its reach. Users should combine secure practices and endpoint defenses rather than rely solely on built-in protections.

🔒 Chrome Enterprise Premium extends the familiar Chrome browser with enterprise-grade security, management, and healthcare-focused integrations to help clinicians access EHRs and web apps securely. It highlights advanced DLP, real-time URL scanning, phishing and malware protections, and controls to block copy/paste, printing, or screen capture of sensitive records. Google announced partnerships with Epic, Imprivata, AuthX, and Citrix to optimize EHR performance, enable passwordless authentication, streamline identity, and unify virtualized workflows. Eligible organizations are offered an extended 6-month trial, no-charge seats for the trial, $5,000 in services funding upon conversion, and high-touch onboarding.

🛡️ Unit 42 identified 18 malicious browser extensions posing as GenAI productivity tools that deliver RATs, infostealers and MitM capabilities. These extensions intercept prompts, exfiltrate credentials and proxy HTTPS responses, often using AI-generated code to accelerate development. Organizations should restrict extensions, scrutinize permissions and treat browsers as critical attack surfaces. Google removed or warned developers after disclosure.

🔍 A LayerX Security study found more than 80 widely used browser extensions explicitly reserve the right to collect and sell user data, with millions of combined installations across categories such as streaming, ad blocking and productivity. The researchers reported that 71% of Chrome Web Store extensions do not publish a privacy policy, leaving many users without visibility into how their data is handled. The findings detail networks of media extensions aggregating viewing behavior and at least a dozen ad blockers and 29 business-focused extensions that may expose enterprise browsing activity. LayerX recommends organisations adopt centralized extension governance and add privacy policy review to extension evaluation criteria.

🔒 Researchers at Socket uncovered 108 malicious Google Chrome extensions that collectively amassed about 20,000 installs and reported to a single command-and-control server. Published under five publisher identities, the add-ons posed as games, Telegram sidebars, and enhancement tools while exfiltrating Google account data, hijacking Telegram Web sessions, opening arbitrary URLs, and injecting ads and scripts. Some source files contained Russian-language comments; attribution remains unconfirmed. Users should remove any identified extensions and log out of Telegram Web sessions immediately.

🔔 Researchers uncovered 'Pushpaganda', an ad fraud campaign that uses search engine poisoning and AI-generated content to surface deceptive stories in Google Discover and trick Android and Chrome users into enabling persistent browser notifications. Once enabled, the alerts deliver scareware-style legal threats and redirect victims through actor-controlled domains that generate illicit ad revenue and funnel users to financial scams. HUMAN's findings link the operation to hundreds of domains and hundreds of millions of bid requests, and Google has deployed a fix.

🚨Research by Socket uncovered a coordinated campaign of 108 malicious Chrome extensions that affected about 20,000 users. Distributed across gaming, social media and translation categories, these extensions appear legitimate while quietly harvesting sensitive data, including Google profiles and active web sessions. Operators used a single command-and-control infrastructure and shared code, complicating detection and enabling a Malware-as-a-Service model.

🔒 A new report from LayerX warns that AI browser extensions are an unmonitored consumption channel that bypasses DLP and SaaS logs, granting direct access to page content, inputs, cookies, and sessions. AI extensions are significantly more likely to contain CVEs and to request scripting, cookie, or tab-manipulation permissions, and they frequently expand privileges after installation. The report urges continuous extension inventories, behavior-based controls, and stricter trust criteria to reduce exposure without hindering productivity.