All news with #vertex ai tag

🚀 Developers using Ray Serve for LLM inference on Google Kubernetes Engine (GKE) now get significantly better performance thanks to a joint effort with Anyscale. Three architectural changes — HAProxy integration for internal routing, a direct token streaming path, and a v2 Ray executor backend for vLLM — reduce overhead and latency. Benchmarks on A4 VMs with NVIDIA HGX B200 hardware show up to 5x higher throughput and 8x lower latency, while preserving Ray's developer-friendly features.

🔒 A design flaw in the Vertex AI SDK for Python allowed attackers to hijack model staging buckets across projects by predicting bucket names derived from project ID and region. Unit 42 researchers called this class of issue Bucket Squatting, where global bucket name uniqueness enabled pre-creation and silent takeover. The flaw could lead to cross-tenant model poisoning and remote code execution via pickle deserialization. Google issued fixes in SDK versions 1.144.0 and 1.148.0 and users should upgrade.

🛡️ Palo Alto Networks Unit 42 disclosed a flaw in the Google Cloud Vertex AI Python SDK that let an attacker with only their own Google Cloud project and a victim's project ID hijack model uploads and execute code in Vertex AI serving containers. Google fixed the issue; users must update to google-cloud-aiplatform version 1.148.0 or later and explicitly set a staging_bucket. The bug arose from predictable default bucket names and lack of ownership checks, enabling an attacker to precreate the bucket, swap uploaded model files (often pickled), and run malicious code when Vertex AI loaded the model.

🛡️ We discovered a vulnerability in the Google Cloud Vertex AI Python SDK that allowed an attacker to hijack a model upload and poison it, enabling remote code execution (RCE) in a victim's serving infrastructure. The issue stems from a predictable default staging bucket name and a missing ownership check in the SDK. By creating the same deterministic bucket in their own project and granting broad permissions, an attacker could replace uploaded model artifacts within a short window before Vertex AI reads them. Google fixed the issue in google-cloud-aiplatform v1.148.0 released April 15, 2026; developers should upgrade to the patched SDK.

🦷 Movix built a custom agentic AI platform to address a severe shortage of skilled dental technicians and reduce costly remakes in aligner and appliance manufacturing. Using Google Cloud infrastructure, including Gemini Enterprise Agent Platform, Cloud Run with L4 GPUs, and Compute Engine, Movix developed deep learning, computer vision, and 3D mesh models to automate quality control and data entry. The solution integrates with legacy lab systems, anonymizes PHI for compliance, and targets large-volume labs to improve accuracy, speed, and cost savings.

🛠️ At Google I/O 2026, Google announced expanded integration between AI Studio and Google Cloud, allowing new users to deploy up to two full-stack apps on the Starter Tier without a billing account. Developers can now choose between Firestore (non-relational) and Cloud SQL (relational) with Firebase Auth for unified authentication. The AI agent can infer or provision the appropriate database, provision resources, generate schema and code, and deploy apps directly to Cloud Run for rapid prototyping.

🤖 At Google Cloud Next, public sector leaders showcased how they are using AI agents to boost productivity and mission impact across government and research organizations. Google introduced the Gemini Enterprise Agent Platform—an evolution of Vertex AI—plus the Gemini Enterprise App with Gemini 3.1 Pro and an Agent Designer for inspectable, schedule‑based workflows. The announcement also covered AI infrastructure (TPU 8 series), an Agentic Data Cloud, enhanced security and Agentic Defense, partner initiatives, and upskilling through the GEAR program.

🚀 At Google Cloud Next ’26, Google presented a unified stack to move AI into enterprise production, anchored by Gemini Enterprise as the connective tissue between data, people, and goals. Key launches include the Gemini Enterprise Agent Platform for building, scaling, governing, and optimizing agents, and the AI Hypercomputer with next-generation TPU 8 chips. Google also outlined the Agentic Data Cloud to ground agents in enterprise context, expanded security agents in Agentic Defense, Workspace Intelligence enhancements, and cross-cloud data capabilities to accelerate real-world deployment.

🚀 Google Cloud has integrated partner-built agents from its Agent Marketplace into the Agent Gallery inside the Gemini Enterprise app, creating a centrally governed hub for discovering and managing specialist, role-specific AI. Featured partners — including Accenture, Adobe, Atlassian, Palo Alto Networks, Salesforce and others — must pass a four-step evaluation to earn the Google Cloud Ready - Gemini Enterprise badge. Built-in safeguards such as cryptographic agent identities, Agent Gateway, and Model Armor protect data and prevent use for model training. Customers can trial the Gallery, while partners can apply to the AI Agents Program and access a rapid deployment framework.

🚀 Gemini Enterprise is an end-to-end system for the agentic era, combining access to frontier models, a developer platform, a collaborative app, and a partner ecosystem to build and deploy agent fleets. The offering centers on the Gemini Enterprise Agent Platform — an evolution of Vertex AI — with an enhanced Agent Development Kit (ADK), graph-based orchestration, persistent Memory Bank, and fast Agent Runtime for multi-step work. IT teams gain a unified control plane for identity, governance, Model Armor, and auditing, while knowledge workers use a no-code Agent Designer, Inbox, Projects, and Canvas to create and monitor agents.

📣 This article describes how the Dev Signal team transitioned a multi-agent prototype into production on Google Cloud by combining a FastAPI service, a Vertex AI memory bank, and the Agent Developer Kit. It highlights production-ready concerns including OpenTelemetry traces exported to Cloud Trace for visibility into agent reasoning, and secure secret handling via Secret Manager so credentials never appear in environment variables. The guide also demonstrates reproducible infrastructure using Terraform to provision Artifact Registry, service accounts, Cloud Run, and related APIs, and outlines containerization and Cloud Build steps to deploy new revisions.

🟢 Claude Opus 4.7 is now generally available on Vertex AI, delivering improved problem solving, instruction following, and expanded vision and long-memory capabilities. The release boosts accuracy on high-resolution documents and charts and enhances performance in coding and agentic workflows. Paired with Vertex AI’s infrastructure, you can scale agents, leverage low latency and provisioned throughput, and apply unified security controls and Model Armor. Access is available on Vertex AI and via Google Cloud Marketplace with sample notebooks and pricing guidance.

⚡ This post outlines an event-driven architecture that pairs BigQuery continuous queries with Pub/Sub Single Message Transforms and ADK-powered agents on Vertex AI Agent Engine to detect, route, and resolve anomalies in real time. Continuous queries push precise, filtered events into Pub/Sub where SMTs reshape payloads for agent webhooks. Deployed agents investigate autonomously, escalate complex cases, and log analytics back into BigQuery for observability.

🤖 In collaboration with Google DeepMind, the team built an AI pose-estimation pipeline that converts single 2D video into a 63-joint 3D biomechanical model for U.S. Olympians. The system uses learned temporal priors to infer occluded joints and delivers near-instant results by running models on statically provisioned TPU slices. Orchestration, scaling, and security are managed with Vertex AI and VPC private endpoints.

🧪 This article describes how to validate the Dev Signal multi-agent system locally before deploying to Cloud Run. It covers configuring local secrets, an environment-aware env utility that initializes Vertex AI, and a test runner which connects to the cloud-based Vertex AI memory bank to persist user preferences. The guide demonstrates a two-phase scenario that teaches preferences, generates multimodal content, wipes local session history, and verifies cross-session memory recall.

🔒 Anthropic’s newest and most capable model, Claude Mythos Preview, is available in Private Preview to a select group of Google Cloud customers through Project Glasswing. Its placement on Vertex AI provides enterprises access to a frontier model integrated with Google Cloud’s tools to build, scale, and govern AI applications and agents. The announcement emphasizes high performance across use cases and a renewed focus on reducing cybersecurity risk in enterprise deployments.

🏠 Rightmove migrated from siloed on-premises databases to Google Cloud to build a unified analytics and AI platform it calls the data hive. Using BigQuery, Vertex AI, and Looker, the company extracts metadata from listings and images to deliver personalized search, agent-assist messaging, and an Automated Valuation Model. The hub-and-spoke architecture centralizes governance while enabling business units to run tailored forecasting and ML use cases. Around 300 staff now use the platform to convert data into operational and commercial value.

🎵 This guide outlines best practices for prompting Lyria 3 and Lyria 3 Pro, Google’s music generation models that deliver granular control over vocals, instrumentation, arrangement, and timing. It highlights technical details—track lengths from rapid 30-second prototypes to three‑minute compositions, multi‑vocal support in eight languages, timed-lyrics and tempo conditioning—and includes a concise prompting framework. The post also covers advanced workflows such as timestamped segment instructions and multimodal generation using images or PDFs, plus integration paths through Vertex AI and the Gen AI SDK.

🎵 Google has made Lyria 3 and Lyria 3 Pro available on Vertex AI in public preview, bringing high-fidelity music generation to the Vertex AI API and Media Studio. Lyria 3 Pro composes studio-quality tracks up to three minutes with structural elements (intros, verses, choruses, bridges), while Lyria 3 produces 30-second tracks for rapid prototyping. Both models accept multi-modal inputs (text or images), support vocal generation with timed lyrics or user-provided lyrics, and can produce purely instrumental pieces. Outputs are embedded with SynthID watermarking and filtered for policy and IP compliance.

🚀 Google Cloud has launched Veo 3.1 Lite, a cost‑effective video generation model available now on Vertex AI, and introduced a new standalone Veo upscaling capability currently in private preview. The Veo 3.1 family now includes three tiers—Veo 3.1, Veo 3.1 Fast, and Veo 3.1 Lite—all with native audio generation. The upscaling tool enhances existing low‑resolution videos to 1080p and 4K, regardless of source, and access is provided via the Vertex AI API and Vertex AI Media Studio. Developer documentation and a sample video editor agent are available to help teams get started.