All news with #google tag

⚠️ A short experiment demonstrates how easy it is to poison AI outputs by publishing a single fabricated webpage. The author wrote an entirely false article titled "The best tech journalists at eating hot dogs," inventing events and rankings; within 24 hours Google Gemini and ChatGPT had incorporated the falsehoods, while Claude resisted. The incident underscores the fragility of trust in AI-derived answers.

🕒 Firefly is a software-driven clock synchronization system from Google that achieves nanosecond-level timing across data center NICs using commodity hardware. It separates fast internal NIC-to-NIC consensus from external UTC alignment and builds consensus over a d-regular random graph. Practical techniques—RTT filtering, path profiling, and optional switch/NIC features—reduce jitter and asymmetry. It yields consistent sub-10ns internal alignment while scaling to large fabrics.

🔒 Three former Google engineers and one spouse were indicted in U.S. federal court for allegedly stealing trade secrets and transferring sensitive files, including materials related to Google's Tensor processor, to unauthorized locations reportedly including Iran. The defendants — Samaneh Ghandali, Mohammadjavad Khosravi and Soroor Ghandali — are accused of exfiltrating documents to third‑party channels, copying files to personal and employer devices, and concealing their actions. They were arrested in San Jose after Google detected suspicious activity and notified law enforcement; the indictment carries multiple counts with significant prison and fine exposures.

🛡️ ESET researcher Lukas Stefanko has identified PromptSpy, the first known Android malware to call a generative AI model at runtime, leveraging Google's Gemini to adapt persistence on different devices. The malware submits an XML dump of the current UI plus a chat prompt to Gemini, receives JSON-formatted instructions, and uses the Accessibility Service to pin the app in Recent Apps in a loop until confirmed. Its primary payload is a VNC-based spyware module that can capture PINs, record unlock patterns and screen activity, take screenshots, and report foreground apps. To block removal it overlays invisible UI elements over uninstall or permission controls; victims must reboot into Safe Mode to remove it.

🛡️ ESET researchers disclosed PromptSpy, the first Android malware observed to integrate Google's Gemini generative AI into its execution flow and achieve persistence. The malware assigns Gemini the persona of an 'Android automation assistant,' sends an XML dump of the current screen, and receives JSON step-by-step instructions that are executed via accessibility services. PromptSpy captures lockscreen data, records screens and video, deploys a VNC module for remote access, and blocks uninstallation using invisible overlays while communicating with a hard-coded C2.

🔒 In 2025 Google reports expanded AI-driven, multi-layered defenses across Google Play and Android, preventing 1.75 million policy-violating apps and banning over 80,000 developer accounts. The company integrated generative AI into review workflows, boosted Play Protect scanning to 350 billion app checks daily, and blocked 266 million risky sideload installation attempts. Developer tools, verification, and pre-review checks aim to reduce friction while raising security and privacy standards.

🛡️ Google says it blocked more than 1.75 million apps from being published on Google Play in 2025 and denied over 255,000 apps access to sensitive user data. The company also banned over 80,000 developer accounts and strengthened detection by integrating generative AI into its review process to identify evolving malicious patterns. Play Protect scanned an estimated 350 billion app instances daily and flagged over 27 million malicious sideloaded apps, while the Play Integrity API processed more than 20 billion checks per day.

🔒 In 2025 Google deployed advanced AI-powered defenses across Google Play to stop apps that cause real-world harm, preventing over 1.75 million policy-violating apps from being published and banning more than 80,000 malicious developer accounts. We expanded Google Play Protect to scan over 350 billion Android apps daily and rolled out real-time protections including in-call scam defenses. Together these measures strengthened the Play store and the broader Android ecosystem by prioritizing automated detection, rapid enforcement, and on-device user safety.

☁️ Google Cloud has made the native endpoint for Cassandra Query Language (CQL) generally available on Spanner, allowing Apache Cassandra applications to use familiar CQL with minimal code changes. Customers can migrate with as little as a one-line change and leverage Spanner’s strong consistency, virtually unlimited scale, geo-partitioning, and enterprise SLAs. The release pairs the CQL compatibility layer with bulk and live migration tools so teams can consolidate operations onto a managed, multi-model platform.

🎿 Using Google Cloud AI and Gemini-era vision from DeepMind, the Google Cloud Team and U.S. Ski & Snowboard built a tool that extracts full 3D biomechanical pose data from ordinary video to measure rotation, posture, and airtime. The system defines a torso-based Body Frame and uses quaternions to compute true angular travel, producing a new metric called Rotational Degrees. Visual outputs like the Cork Ribbon reveal axis tilt and consistency for coaches and broadcasters, helping quantify efficiency and the physical feasibility of higher-degree tricks.

🔁 BigQuery now provides autonomous embedding generation that automatically maintains an embedding column derived from a source column, eliminating manual embedding pipelines and synchronization burdens. The managed column is defined via familiar SQL (GENERATED ALWAYS AS AI.EMBED) and can be stored asynchronously. It integrates with VECTOR_SEARCH and a new AI.SEARCH function, and supports secure calls to Vertex AI models with connection-based permissions. The feature is available in preview.

💬 This Google Cloud post by David Tamaki Szajngarten demonstrates how to build a context-aware conversational agent for BigQuery using the Conversational Analytics API powered by Gemini. It provides a reference Python SDK flow to register BigQuery tables, create a DataAgent with system instructions and permitted datasources, and deploy stateful or stateless conversations. The article shows a streaming chat loop that returns generated SQL, DataFrame-like results, Vega‑Lite chart specs, and final natural-language answers, and highlights integration with the Agent Development Kit (ADK) and lifecycle controls.

🔒 Google Cloud argues Europe does not need to choose between digital sovereignty and economic growth. The company promotes a partnership-led model that combines advanced AI capabilities (including Gemma and Gemini-class technology) with flexible infrastructure and multicloud choice. Emphasis is placed on interoperability, local control, and security certifications such as SecNumCloud, along with options like Google Cloud Air-Gapped and Dedicated deployments. The aim is to enable growth while preserving control and resilience.

🚀 Google announced Gemini 3.1 Pro, an upgraded foundation model in the Gemini 3 series that emphasizes deeper reasoning and complex problem solving. The model is available in preview in Vertex AI and Gemini Enterprise, and developers can access it through Google AI Studio, the Gemini API, Android Studio, Google Antigravity, and the Gemini CLI. Early customers report meaningful gains in speed, efficiency, and accuracy across code, 3D transformations, and product design workflows.

🧠 ESET researchers have identified PromptSpy, the first known Android malware to integrate generative AI (Google's Gemini) into its execution flow. The malware sends serialized UI XML to Gemini and receives JSON-formatted tap, swipe, and long-press instructions to navigate device-specific interfaces. This enables robust persistence by programmatically locking the app in Recent Apps and deploying a VNC module for remote control and data exfiltration. Distribution appears limited and regionally focused, but the technique raises broader concerns about AI misuse.

🌐 Google today announced America-India Connect, a multi-year initiative anchored by a five-year $15 billion AI infrastructure investment in India to expand connectivity across four continents. The program will build a new subsea gateway in Visakhapatnam, add three subsea paths linking India to Singapore, South Africa, and Australia, and deploy four strategic fiber-optic routes between the United States, India, and Southern Hemisphere locations. These investments aim to boost network resilience, capacity, and affordable access to digital services while supporting skilling and government training platforms.

🔒 At the 62nd Munich Security Conference, Kent Walker (President, Google & Alphabet) argued that fragmented defenses are inadequate against AI-accelerated cyber threats and the near-term risk from cryptographically relevant quantum computing. Google highlighted GTI findings that adversaries are automating reconnaissance and producing hyper-realistic phishing, and showcased the Ukrainian startup LetsData, which uses AI to scan multilingual media and detect InfoOps at scale. To scale defender advantages, Google launched the Gemini Startup Forum: Cybersecurity and promotes deployment options such as Google Distributed Cloud Air-Gapped for sovereign, secure use of its infrastructure. Walker urged governments, industry, and vendors to adopt a full-stack, collaborative approach—breaking silos and modernizing procurement—to build shared digital resilience.

🌐 BigQuery now supports global queries (preview), letting you run a single standard SQL statement across datasets stored in different geographic regions without building ETL pipelines. BigQuery orchestrates distributed execution by running region-specific operations, transferring partial results to a chosen execution location, and combining them to produce final results. The feature is disabled by default and requires administrator opt-in and explicit per-user permissions. It respects VPC Service Controls and gives you control over where queries execute to help meet compliance and data residency requirements.

🔒 Google has released an urgent security update for Chrome to address CVE-2026-2441, a high-severity zero-day affecting desktop builds on Windows, macOS and Linux. The flaw, rooted in a CSS processing issue, can allow a remote attacker to execute arbitrary code inside a sandbox via a crafted HTML page. Google confirmed an exploit is already in the wild and credited researcher Shaheen Fazim for reporting the bug on February 11; the company issued the patch on February 13.

🔧 Google released emergency updates to fix a high-severity Chrome zero-day (CVE-2026-2441) that is being exploited in the wild. The flaw is a use-after-free caused by an iterator invalidation bug in CSSFontFeatureValuesMap, and Google pushed a backported patch across stable branches. Fixes are rolling out to Windows and macOS (145.0.7632.75/76) and Linux (144.0.7559.75); users should update or let Chrome apply updates automatically. Google noted additional related work remains tracked in bug 483936078.