All news with #google tag

Google: AI Studio Aims to Let Everyone 'Vibe Code' Games

🕹️ Google says its AI Studio will enable users to 'vibe code' simple video games by the end of the year. The company claims the tool can automatically select models and wire up APIs to streamline app creation, while noting current limitations for production-ready systems. Product lead Logan Kilpatrick highlighted the potential to broaden access to game creation, and startups like Cursor are pursuing similar next-generation vibe coding tools.

How Five Agencies Built Impossible Ads with Gemini

🎨 Google showcased how five agencies used Gemini 2.5 Pro and complementary generative media models to produce ambitious ad campaigns that blend nostalgia, personalization, and scalable visual storytelling. Projects ranged from a retro AI radio for Slice to personalized "postcard" ads for Virgin Voyages, AI co-hosts and party themes for Smirnoff, crowdsourced mascots for Visit Orlando, and cinematic short film work with Moncler. Results highlighted rapid production, measurable engagement lifts, and cross-product workflows across Imagen, Veo, Lyria, and Vertex AI. The post invites brands to explore these tools for creative scale and efficiency.

Proteomics AI Agent: Guided Protocols and Error Detection

🔬 Researchers at the Max Planck Institute of Biochemistry and Google Cloud created a Proteomics Lab Agent using the Agent Development Kit and Gemini models to provide personalized, multimodal AI guidance for mass spectrometry experiments. The agent analyzes recorded steps to generate publication-ready protocols, detect procedural errors, and capture tacit expertise into a searchable knowledge base. Open-sourced on GitHub, it aims to reduce troubleshooting time and improve reproducibility across labs.

Google Cloud Cross-Site Interconnect Now Generally Available

🔗 Google Cloud announced that Cross-Site Interconnect is now generally available, providing a cloud-delivered, transparent layer 2 connectivity overlay for global WANs. The service uses Google’s global backbone to deliver high-bandwidth, on-demand links with a 99.95% SLA, dynamic consumption pricing, and support for MACsec encryption and VLAN-based multi-destination handoffs. It exposes operational metrics via dashboards and APIs so teams can monitor latency, loss, and utilization while abstracting failover and resiliency management.

Vietnam Actors Use Fake Job Postings to Hijack Ad Accounts

🔎 GTIG describes a targeted campaign by a Vietnam-based cluster tracked as UNC6229 that uses fake job postings on legitimate platforms to socially engineer remote digital advertising workers. Victims are enticed to open password-protected attachments or visit convincing phishing portals that harvest corporate credentials and can bypass MFA. The actors abuse reputable CRM and SaaS services to increase trust, deliver remote access trojans, and ultimately take over high-value advertising and social media accounts for sale or resale.

Google Gen AI .NET SDK Brings Gemini to C#/.NET Developers

🚀 Google has released the Google Gen AI .NET SDK, bringing unified access to Gemini on Google AI and Vertex AI for C#/.NET developers. The SDK is available via NuGet (dotnet add package Google.GenAI) and supports client creation with an API key or with project/location settings for Vertex AI. Examples demonstrate unary and streaming text generation, image generation, and configurable response schemas and generation settings. Google provides the API reference, GitHub source (googleapis/dotnet-genai) and a DemoApp with samples to help developers get started.

YouTube Ghost Network: Disrupting a Massive Malware Campaign

🛡️ Check Point Research uncovered the YouTube Ghost Network, a large-scale operation that used fake and compromised accounts to distribute infostealers like Rhadamanthys and Lumma. More than 3,000 malicious videos — often disguised as cracked software or game hacks — were reported and removed after being linked to password-protected archives that carried the malware. Compromised accounts, coordinated comment manipulation, and false endorsements were used to build trust and drive downloads.

Google Cloud H4D VMs Boost Finance Workload Performance

⚡ Google Cloud announced the H4D VM family (Preview), powered by 5th Gen AMD EPYC processors (Turin), aimed at delivering extreme performance for financial services workloads. The H4D series targets latency-sensitive use cases such as high-frequency trading, Monte Carlo risk simulations, backtesting, and derivatives pricing by offering faster core-to-core communication, larger memory capacity, and improved network throughput. AMD benchmarking with the open-source KX Nano test reported an average ~34% out-of-the-box performance gain over prior C3D VMs, with per-core and multi-threaded uplifts around 1.33–1.36x. Google Cloud will demonstrate H4D and complementary HPC solutions at STAC Summit NYC on October 28th and will have experts available to discuss performance, security, and compliance.

Model Armor and Apigee: Protecting Generative AI Apps

🔒 Google Cloud’s Model Armor integrates with Apigee to screen prompts, responses, and agent interactions, helping organizations mitigate prompt injection, jailbreaks, sensitive data exposure, malicious links, and harmful content. The model‑agnostic, cloud‑agnostic service supports REST APIs and inline integrations with Apigee, Vertex AI, Agentspace, and network service extensions. The article provides step‑by‑step setup: enable the API, create templates, assign service account roles, add SanitizeUserPrompt and SanitizeModelResponse policies to Apigee proxies, and review findings in the AI Protection dashboard.

Google Migrates ISAs with AI and Automation at Scale

🔧 Google details how its custom Axion Arm CPUs and a mix of automation and AI enabled large-scale migration from x86 to multi-architecture production across services such as YouTube, Gmail, and BigQuery. The team analyzed 38,156 commits (about 700K changed lines) and reports migrating more than 30,000 applications to Arm while keeping both Arm and x86 in production. Existing automation like Rosie, sanitizers, fuzzers, and the CHAMP rollout framework handled much of the work, while an LLM-driven agent called CogniPort fixed build and test failures, showing a 30% success rate on a 245-commit benchmark. Google plans to default new apps to multiarch and continue refining AI tools to address the remaining long tail.

Dataplex Supports Column-Level Lineage for BigQuery

🔍 Dataplex Universal Catalog now captures column-level lineage for BigQuery, extending object-level tracing to granular column transformations at no extra cost. The update provides interactive visual lineage graphs so users can inspect upstream and downstream flows for individual columns, trace origins, and assess downstream impact of modifications. This granularity helps validate authoritative sources for AI/ML features, enforce column-level governance, and improve compliance. It also surfaces freshness and usage metadata to support context-aware agents.

SmarterX Builds Custom LLMs with Google Cloud Tools

🔍 SmarterX uses Google Cloud to build custom LLMs that help retailers, manufacturers, and logistics companies manage regulatory compliance across product lifecycles. Using BigQuery, Cloud Storage, Gemini, and Vertex AI, the company ingests, normalizes, and indexes unstructured regulatory and product data, applies RAG and grounding, and trains customer-specific models. The integrated platform empowers subject matter experts to evaluate, correct, and deploy model updates without heavy engineering overhead.

Cloud Armor: Hierarchical Policies, Extended WAF and NTI

🛡️ Cloud Armor introduces hierarchical security policies and organization-scoped address groups to simplify centralized policy management across organization, folder, and project levels. The release also includes GA support for JA4 network fingerprinting and ASN/NTI controls for Media CDN, while an enhanced WAF request-body inspection (preview) expands inspection from 8 KB to 64 KB. These updates are designed to strengthen threat protection and reduce operational complexity for hybrid and multicloud deployments.

Sophisticated Investment Scam Impersonates Singapore Official

🔍 Cybersecurity researchers have uncovered a large-scale investment scam that impersonated Singapore’s top officials, including Prime Minister Lawrence Wong and Minister K Shanmugam, to promote a fraudulent forex platform. The campaign used verified Google Ads, hundreds of fake news domains and deepfake videos, funneling victims through multiple redirects to a Mauritius-registered trading site. Group-IB reported advanced evasion techniques and localized targeting to show scam pages only to Singaporean users, pressuring many to invest and then blocking withdrawals.

The AI Fix #73: Gemini gambling, poisoning LLMs and fallout

🧠 In episode 73 of The AI Fix, hosts Graham Cluley and Mark Stockley explore a sweep of recent AI developments, from the rise of AI-generated content to high-profile figures relying on chatbots. They discuss research suggesting Google Gemini exhibits behaviours resembling pathological gambling and report on a Gemma-style model uncovering a potential cancer therapy pathway. The show also highlights legal and security concerns— including a lawyer criticised for repeated AI use, generals consulting chatbots, and techniques for poisoning LLMs with only a few malicious samples.

Pro-Russia Information Operations After Drone Incursion

🔎 Google Threat Intelligence Group (GTIG) observed coordinated pro-Russia information operations responding to reported Russian drone incursions into Polish airspace on Sept. 9–10, 2025. Actors amplified narratives denying Russian culpability, blaming NATO or Poland, and seeking to erode domestic and international support for Ukraine. GTIG documented activity across multiple networks and languages and noted these operations leveraged both long-standing and recently developed influence infrastructure.

Google abandons Privacy Sandbox, ends most cookie efforts

🍪 Google has announced it is discontinuing 11 Privacy Sandbox technologies — effectively ending most of the company’s cookie‑replacement efforts after evaluating low adoption and ecosystem feedback. The decision follows regulatory scrutiny from the UK’s Competition and Market Authority and several U.S. antitrust actions, and came after prior concessions from Google. The company says it will continue to work on privacy improvements for Chrome, Android and the web but will move away from the Privacy Sandbox branding.

Coldriver Deploys New 'NoRobot' Malware Suite, 2025

🛡️ Google Threat Intelligence Group (GTIG) has observed the Russian-linked Coldriver group deploying a new, staged malware ecosystem tracked as NoRobot, YesRobot and MaybeRobot. GTIG's October 20, 2025 report shows the campaign replaces the previously disclosed LostKeys strain and begins with a 'ClickFix-style' ColdCopy phishing lure that tricks victims into running a malicious DLL via rundll32.exe. NoRobot functions as a downloader using split-key cryptography and staged payloads; operators briefly used a Python-based backdoor (YesRobot) before switching to a more flexible PowerShell backdoor (MaybeRobot) to reduce detection.

DNS0.EU DNS Service Shuts Down Over Sustainability Concerns

🔒 The DNS0.EU non‑profit public DNS resolver announced an immediate shutdown, citing unsustainable time and resource constraints for its volunteer team. Launched in 2023 and operated from France with 62 servers across 27 cities in all EU member states, the service supported no‑logs policies and modern encrypted transports including DNS‑over‑HTTPS, DNS‑over‑TLS, and DNS‑over‑QUIC. The operators thanked partners and urged users to migrate to DNS4EU or NextDNS, both of which offer privacy protections and defenses against malicious domains.

G4 VMs: High-performance P2P Fabric for Multi‑GPU Workloads

🚀 Google Cloud's newly GA G4 VMs combine NVIDIA RTX PRO 6000 Blackwell GPUs with a custom, software-defined PCIe fabric to enable high-performance peer-to-peer (P2P) GPU communication. The platform accelerates collective operations like All-Gather and All-Reduce without code changes, delivering up to 2.2x faster collectives. For tensor-parallel inference, customers can see up to 168% higher throughput and up to 41% lower inter-token latency. G4 integrates with GKE Inference Gateway for horizontal scaling and production deployments.