All news with #google tag

🚨 A Google Chrome extension carrying a "Featured" badge, Urban VPN Proxy, has been found silently harvesting prompts and responses from major AI chat services and sending them to remote analytics servers. The extension — installed by roughly six million Chrome users and about 1.3 million Edge users — was updated on July 9, 2025 (v5.5.0) with AI capture enabled by default. Injected scripts override browser networking APIs to intercept chat data and exfiltrate conversation text, IDs, timestamps, session metadata, and model/platform information. The publisher's updated privacy policy admits collecting AI prompts and outputs for "Safe Browsing" and marketing while disclaiming a full guarantee of de-identification.

📊 Cloudflare’s Radar report summarizes the Top Internet Services of 2025 using anonymized DNS queries from the 1.1.1.1 resolver and a machine-learning ranking method. It highlights continued dominance by Google and Facebook, strong gains by generative AI like ChatGPT and emerging rivals, and regional shifts such as Kwai rising in emerging markets. The analysis spans nine categories and includes country-level Top 10s for local context. E-commerce momentum saw Shopee and Temu join Amazon in the global top three, while crypto, news, and streaming showed event-driven volatility.

🔒 Google's Threat Intelligence Group linked five additional China-aligned cyber-espionage groups to active exploitation of the maximum-severity CVE-2025-55182 React2Shell remote code execution flaw affecting React and Next.js server components. Attackers are executing commands and exfiltrating AWS configuration files and credentials from vulnerable hosts; Palo Alto and AWS reported widespread breaches. Shadowserver and GreyNoise are tracking tens of thousands of exposed systems and hundreds of exploit attempts. Organizations should urgently patch affected React 19.0–19.2.0 releases and apply mitigations.

🔊 The post announces general availability of Gemini Live API on Vertex AI, powered by the Gemini 2.5 Flash Native Audio model. It presents a unified, low-latency native audio architecture that replaces multi-stage STT/LLM/TTS pipelines and enables real-time multimodal reasoning over audio, text, and visual streams via a stateful WebSocket. Two quickstart templates (Vanilla JS and React) and three production demos illustrate common integration patterns, partner telephony/WebRTC support, and recommended backend proxying for secure credentials.

🔗 This post explains how to expose Looker’s semantic layer to Gemini Enterprise quickly by using the MCP Toolbox for Databases and the Agent Development Kit (ADK). It outlines three concise steps: deploy the MCP Toolbox (recommended to Cloud Run), build and deploy an ADK agent to Vertex AI Agent Engine, and register that agent with Gemini Enterprise. The result: trusted Looker models available inside Gemini for natural‑language business queries.

🔒 Google is offering public sector teams a no-cost 30-day test drive of Google Workspace, providing full access to its cloud-native productivity suite with embedded Gemini AI. The pilot includes guided setup, user onboarding, workflow testing, and an impact assessment report to quantify productivity and security benefits. Workspace with Gemini has FedRAMP High authorization and Google Public Sector cites CMMC Level 2 validation for internal systems handling CUI. Agencies can evaluate real workflows commitment-free and gather data to inform adoption decisions.

🔹Google's Data Cloud presents an AI-native approach to unify siloed, multimodal enterprise data and enable real-time insights. The platform integrates Gemini with BigQuery to automate pipeline building, vector embedding, and data quality, while Dataplex provides continuous cataloging and contextual metadata for agents. Enhancements to managed streaming, Pub/Sub UDFs, and Dataflow extend real-time processing, governance, and developer productivity.

🔊 Gemini Live API, powered by the Gemini 2.5 Flash Native Audio model, is now generally available on Vertex AI. It enables low-latency, multimodal conversational agents that combine voice, vision, and text to deliver human-like, contextual interactions. The API supports natural turn-taking, acoustic cue analysis, and visual understanding, and is optimized for enterprise-scale, regional deployments and compliance. Early adopters including Shopify, United Wholesale Mortgage, and SightCall report improved efficiency and real-time assistance.

⚠ CISA added CVE-2025-14174, a Google Chromium out-of-bounds memory access vulnerability, to the Known Exploited Vulnerabilities (KEV) Catalog following evidence of active exploitation. This class of flaw frequently enables memory corruption and can lead to code execution or information disclosure, posing significant risk to the federal enterprise. Under BOD 22-01, Federal Civilian Executive Branch agencies must remediate KEV entries by required due dates; CISA urges all organizations to prioritize timely remediation as part of their vulnerability management.

🤖 Offensive AI is accelerating cyberattacks, automating reconnaissance, malware orchestration, and large-scale credential harvesting. Security teams and observers such as Google Threat Intelligence and Anthropic warn these techniques can bypass legacy defenses and overwhelm manual SOC processes. Network Detection and Response (NDR) solutions provide continuous network visibility, real‑time analytics, and automated triage to detect polymorphic threats, reduce false positives, and speed incident response.

🔍 Elastic Security Labs has detailed a Windows backdoor named NANOREMOTE that leverages the Google Drive API to stage payloads and exfiltrate data, making detection more difficult. The C++ implant implements a robust task manager for queued uploads and downloads with pause, resume and cancel capabilities and exposes 22 command handlers for reconnaissance, execution and file transfer. Researchers also observed a WMLOADER dropper and an uploaded artifact linking NANOREMOTE to the FINALDRAFT family, indicating likely code reuse.

🔒 Google released a Chrome security update on December 10 to patch three zero-day vulnerabilities, including a high-severity bug tracked internally as 466192044 for which an exploit is reported in the wild. Google has not published technical details and marks the issue as Under coordination, saying details may be restricted until most users are updated. The advisory also fixes two additional issues: CVE-2025-14372, a use-after-free in Chrome's Password Manager reported by Weipeng Jiang, and CVE-2025-14373, an inappropriate implementation in the Chrome toolbar reported by Khalil Zhani.

🔔 Google has issued emergency updates for Chrome to address a zero-day tracked as Chromium bug 466192044 that is actively exploited in the wild. The vulnerability is a buffer overflow in the LibANGLE Metal renderer caused by improper buffer sizing and can lead to memory corruption, crashes, sensitive data leaks, or arbitrary code execution. Stable channel builds rolling out are Windows 143.0.7499.109, macOS 143.0.7499.110, and Linux 143.0.7499.109; users should update immediately or allow Chrome to install the update on restart.

🔐 Google released Chrome security updates addressing three vulnerabilities, including a high-severity flaw that is being actively exploited in the wild and is tracked as Chromium issue 466192044. Google withheld the CVE identifier, affected component, and technical details while coordinating disclosure to allow broader patching. The release also corrects two medium-severity issues in the Password Manager and Toolbar. Users should update to Chrome 143.0.7499.109/.110 (Windows/macOS) or 143.0.7499.109 (Linux) and apply vendor patches for other Chromium-based browsers when available.

🔒 Google’s Apigee now supports MCP with fully managed, remote servers, enabling organizations to expose existing APIs as agent tools without code changes or running MCP infrastructure. By creating an MCP proxy with your OpenAPI spec and a /mcp basepath, Apigee handles transcoding, protocol handling, and automatic registration in API hub. You can apply Apigee’s built-in security, identity, quota, and analytics controls to govern and monitor agent interactions. The capability is currently available in preview for a limited set of customers.

🔌 Google announced fully-managed, remote support for Anthropic's Model Context Protocol (MCP), enabling agents and standard MCP clients to access a unified, enterprise-ready endpoint for Google and Google Cloud services. The managed MCP servers integrate with services like Google Maps, BigQuery, GCE, and GKE to let agents perform geospatial queries, in-place analytics, and infrastructure operations. Built-in discovery, governance, IAM controls, audit logging, and Google Cloud Model Armor provide security and observability. Developers can expose and govern APIs via Apigee and the Cloud API Registry to create discoverable tools for agentic workflows.

🔒 Google has patched a zero-click vulnerability in Gemini Enterprise and Vertex AI Search that could have allowed attackers to exfiltrate corporate data via hidden instructions embedded in shared Workspace content. Discovered by Noma Security in June 2025 and dubbed "GeminiJack," the flaw exploited Retrieval-Augmented Generation (RAG) retrieval to execute indirect prompt injection without any user interaction. Google updated how the systems interact, separated Vertex AI Search from Gemini Enterprise, and changed retrieval and indexing workflows to mitigate the issue.

🧭 This episode of The Agent Factory unpacks Google's latest AI stack—Gemini 3, the Antigravity IDE, and Nano Banana Pro—through hands-on demos and developer commentary. Guests demonstrate end-to-end workflows, from generating a React Native cataloging app to refactoring a site from screenshots and producing game assets with grounded search. The recap emphasizes enhanced tool use, multimodal inputs, and smoother deployment to Google Cloud. It also highlights the new Vending Bench metric for agentic, long-range decision-making.

🔒 Google’s Android Security and Privacy team collaborated with Arm to analyze the Mali GPU driver and implement SELinux-based IOCTL filtering that reduces the kernel driver's attack surface. The team categorized IOCTLs as unprivileged, instrumentation, and restricted, and used a staged rollout—first opt-in testing via a gpu_harden attribute, then opt-out with a gpu_debug domain—to validate behavior in real devices. The post provides step-by-step guidance for vendors to adopt a platform-level macro, define device-specific IOCTL lists, and enforce policy to keep deprecated and debug IOCTLs unreachable in production.

🔬 AlphaEvolve is a Gemini-powered coding agent on Google Cloud that automates evolutionary optimization of algorithms for complex, code-defined problems. It takes a problem specification, evaluation logic, and a compile-ready seed program, then uses Gemini models to propose mutated code variants and an evolutionary framework to select and refine the best candidates. Early internal results at Google demonstrate measurable efficiency improvements, and the AlphaEvolve Service API is available through a private Early Access Program for interested organizations.