All news with #google tag

🔎 Google is testing AI-powered, interactive UI answers within AI Mode, integrating Gemini 3 to generate on-the-fly interfaces tailored to queries. Instead of relying solely on text and a couple of links, Search can produce dynamic tools—such as an RNA polymerase simulator—to demonstrate concepts in action. This change could improve comprehension but may also reduce traffic to original sites and reshape the web economy.

⚠️CISA has added CVE-2025-13223, a Google Chromium V8 type confusion vulnerability, to its Known Exploited Vulnerabilities (KEV) Catalog following evidence of active exploitation. This class of flaw is a frequent attack vector and poses significant risk to the federal enterprise and other organizations using Chromium-based engines. Under BOD 22-01, Federal Civilian Executive Branch agencies must remediate KEV entries by the required due date; CISA strongly urges all organizations to prioritize timely patching and vulnerability management to reduce exposure.

⚠️ Google released an emergency patch for a high‑severity Type Confusion vulnerability in the V8 JavaScript engine (CVE-2025-13223), which the company says is being exploited in the wild. The flaw, rated CVSS 8.8 and discovered by Clément Lecigne of Google TAG, affects Chromium‑based browsers and can enable heap corruption and potential code execution. Administrators should prioritize updating Chrome to the patched 142.0.7444.175/.176 builds. A second V8 issue, CVE-2025-13224, is also fixed.

🚀 Google Cloud today announced a 30-day free trial for Cloud SQL, offering preconfigured MySQL and PostgreSQL instances to evaluate enterprise-grade features without upfront commitment. The trial exposes Performance, High Availability, and Data Cache capabilities and mirrors production-oriented SKUs (Enterprise and Enterprise Plus). Users can one-click upgrade to paid instances at any time; after 30 days instances are suspended and kept stopped for an additional 90 days at no charge. The offer is available in all Google Cloud regions.

🧬 This guide demonstrates step-by-step fine-tuning of MedGemma (a Gemma 3 variant) to classify breast histopathology images using the public BreakHis dataset and a notebook-based workflow. It highlights practical choices—using an NVIDIA A100 40 GB, switching from FP16 to BF16 to avoid numerical overflows, and employing LoRA adapters for efficient training. The tutorial reports dramatic accuracy gains after merging LoRA adapters and points readers to runnable notebooks for reproducibility.

🔒 Google Cloud's Private NAT enables secure private-to-private translation to connect networks with overlapping or non-routable IPv4 ranges without running NAT appliances. As a managed Cloud NAT feature, it delivers high availability, automatic scalability, and centralized control for hybrid and multi‑VPC topologies. The post includes practical gcloud examples and Network Connectivity Center use cases to guide implementation.

🕒 Google is integrating the TimesFM time-series foundation model into BigQuery and AlloyDB, enabling zero-shot forecasting on customer data without retraining. AI.FORECAST and AI.EVALUATE are now Generally Available in BigQuery, while AI.DETECT_ANOMALIES is in public preview. TimesFM 2.5 offers improved accuracy and lower latency, supports dynamic context windows up to 15K, and can return historical data with forecasts. AlloyDB preview lets users call TimesFM endpoints hosted on Vertex AI so operational data can be forecasted in-place, preserving data residency and reducing export overhead.

🤖 Google has made Gemini 3 available to enterprises and developers via Gemini Enterprise and Vertex AI, bringing advanced multimodal reasoning and agentic capabilities to production teams. The model can analyze text, images, video, audio, and code together, supports a 1M-token context window, and improves frontend generation, legacy code migration, and long-running tool orchestration. Early partners report faster diagnostics, richer UI prototypes, and more reliable automation across business workflows.

🔒 Google released an emergency update to address a newly discovered Chrome zero-day, CVE-2025-13223, which is being actively exploited. The high-severity flaw stems from a type confusion vulnerability in Chrome's V8 JavaScript engine and was reported by Clement Lecigne of Google's Threat Analysis Group. Patches are available in versions 142.0.7444.175/.176 for Windows, 142.0.7444.176 for macOS, and 142.0.7444.175 for Linux; users should check About Google Chrome and relaunch to apply the update.

🛡️ Google has released emergency Chrome updates addressing two V8 engine type confusion flaws, including an actively exploited vulnerability tracked as CVE-2025-13223 (CVSS 8.8) that can lead to arbitrary code execution or crashes. The patch also fixes CVE-2025-13224 flagged by Google's AI agent Big Sleep and completes a set of seven zero-days addressed this year. Users should update Chrome to 142.0.7444.175/.176 (Windows/macOS/Linux) and apply fixes for other Chromium-based browsers when available.

🤖 Google’s Gemini 3 has been spotted in AI Studio, suggesting an imminent rollout that could begin within hours or days. The AI Studio entry references how temperature influences reasoning — noting "For Gemini 3, best results at default 1.0. Lower values may impact reasoning" — and highlights controls such as context size and temperature. Earlier sightings on Vertex AI show a preview build named gemini-3-pro-preview-11-2025, while Google is also testing an image model codenamed GEMPIX2 (Nano Banana 2).

📡 Google announced Dhivaru, a new Trans-Indian Ocean subsea cable linking the Maldives, Christmas Island, and Oman, and will build two new connectivity hubs in the Maldives and Christmas Island. The investment builds on the Australia Connect initiative and aims to improve reach, reliability, and resilience across the Indian Ocean, supporting AI services such as Vertex AI. The hubs will provide cable switching, content caching, and colocation to reduce latency, improve availability, and support local ecosystems.

🔍 This joint analysis from Google Threat Intelligence and Mandiant describes UNC1549 activity observed from late 2023 through 2025 against aerospace, aviation, and defense organizations. The group commonly exploited trusted third‑party relationships, VDI breakouts, and highly targeted spear phishing to gain access, then deployed custom backdoors and tunneling tools to maintain stealth. The report provides IOCs, YARA rules, and detection guidance for Azure and enterprise environments.

🔒 LinkedIn has emerged as a major vector for phishing, with a growing share of attacks moving off email and onto social and messaging platforms. Attackers exploit in‑app DMs, account takeovers, and AI automation to target executives and high‑value roles, often aiming to compromise SSO providers such as Microsoft Entra and Google Workspace. Because these messages bypass traditional email security and lack inbox quarantine tools, browser-based defenses and SSO/MFA hygiene are recommended to detect and block evasive campaigns. The article outlines five reasons this shift increases enterprise risk.

🚀 Google Cloud has launched the Production-Ready AI Learning Path, a free curriculum designed to guide developers from prototype to production. Drawing on an internal playbook, the series pairs Gemini models with production-grade tools like Vertex AI, Google Kubernetes Engine, and Cloud Run. Modules cover LLM app development, open model deployment, agent building, security, RAG, evaluation, and fine-tuning. New modules will be added weekly through mid-December.

🚀 Google Cloud introduces hands-on labs for Gemma 3, a family of lightweight open models offering multimodal (text and image) capabilities and efficient performance on smaller hardware footprints. The labs present two deployment paths: a serverless approach using Cloud Run with GPU support, and a platform approach using GKE for scalable production environments. Choose Cloud Run for simplicity and cost-efficiency or GKE Autopilot for control and robust orchestration to move models from local testing to production.

🔋 Google will begin flagging Android apps on Google Play that show high background activity and cause excessive battery drain. The change centers on a new Android Vitals metric called excessive partial wake locks, and apps that cross the bad-behavior threshold may be labeled as battery drainers and lose prominence in discovery surfaces. Developers will receive alerts in their Android Vitals dashboard and have until March 1, 2026 to remediate issues.

🚀 Google Cloud reports a new state-of-the-art Single Trained Model Track score on the BIRD benchmark, achieving 76.13 with a fine-tuned Gemini 2.5-pro. The team credits rigorous data filtering, multitask supervised fine-tuning, and test-time self-consistency selection for the gains. These improvements bolster NL2SQL features in AlloyDB AI and BigQuery, and enhance developer tooling such as Gemini Code Assist for reliable SQL generation.

🔁 Google has softened its planned Developer Verification requirements after widespread backlash, saying it will create a dedicated account type for limited app distribution and an advanced sideloading flow for experienced users. The original rule would have blocked installation of apps from unverified developers on certified devices beginning in 2026. Google says these changes respond to concerns from students, hobbyists, and power users who need accessible or higher-risk pathways to install apps.

🛡️ Google has filed a civil lawsuit in the Southern District of New York to dismantle Lighthouse, a phishing-as-a-service kit used to power large-scale SMS phishing (smishing) campaigns. The company says the kit — likely run from China and marketed on Telegram — offered more than 600 templates mimicking over 400 organizations and targeted more than one million people across 121 countries. Google is pursuing legal remedies and supporting new legislation while deploying technical protections such as AI-powered scam flagging and expanded account recovery options.