< ciso
brief />
Tag Banner

All news with #google cloud tag

613 articles · page 4 of 31

Google Cloud IAM Helper for Agentic Identity Safety

🔒 Google Cloud introduces IAM Helper, a suite of agent-facing building blocks that automates complex IAM workflows for autonomous agents and CI/CD pipelines. It integrates with core IAM engines like Policy Intelligence and Privileged Access Manager, offering Agent Skills and MCP tools to diagnose permission issues, simulate policy changes, and request just-in-time access while preserving least-privilege.
read more →

LookML semantic indexing now GA in Knowledge Catalog

🔎 Google Cloud announced general availability of LookML semantic language indexing in Knowledge Catalog for Looker Core, enabling enterprise agents to access both organizational context and deterministic BI semantics. The integration automatically indexes LookML business logic into Knowledge Catalog with live synchronization, making metric definitions, join paths, and lineage available to Gemini Enterprise and custom agents. This provides governed, accurate SQL generation and reduces token usage and SQL retry loops while preserving existing governance controls.
read more →

Anthropic-Managed Claude Now Generally Available on Google Cloud

📰 Google Cloud now offers an Anthropic-managed deployment option for Claude models in general availability, alongside the existing Google-managed option. Both provide day-zero access to the same frontier models, identical pricing, and consolidated Google Cloud billing, but differ on who acts as the data processor and timing for new Claude platform features. The Anthropic-managed option runs under Anthropic’s terms with immediate access to platform features, while the Google-managed option provides enhanced governance, VPC Service Controls, CMEK, and regional data residency.
read more →

Knowledge Catalog adds self-learning context enrichment

🛠️ Today Google Cloud previewed self-learning context enrichment and a redesigned Knowledge Studio within Knowledge Catalog. These features use Gemini-powered analysis to detect gaps in organizational context by examining agent and user interactions, then propose verified definitions and relationships drawn from docs, repos, and tables. The tools support governed, reusable context sets, configurable auto-publishing thresholds, and full provenance for changes, aiming to make agents more accurate and efficient.
read more →

Google Cloud launches Agent Compute Engine and governance

🚀 Google Cloud announced Agent Compute Engine, a purpose-built compute service that powers Gemini agents with hardware-isolated microVMs and persistent Hyperdisks, plus Agent Substrate for Kubernetes. The update expands Gemini for business governance across cloud, hybrid, and endpoint runtimes with Agent Gateway, Registry, Identity, and observability features. New previews include cross-cloud enforcement, granular cost controls, model routing, and partner integrations.
read more →

Google Cloud guidance on critical PHP flaws

🛡️ Google Cloud summarizes two critical PHP core vulnerabilities — CVE-2026-109618 and CVE-2026-109619 — disclosed by PHP and Wiz Research that enable unauthenticated remote code execution and memory disclosure. Gemini 4 Argon aided discovery and validation while Google Cloud clarifies it is not directly vulnerable but that customers running PHP may be at risk. The advisory lists affected PHP versions (5.4 through 8.5.11), patched releases, and actionable steps for App Engine, Cloud Run, GKE, and Marketplace deployments.
read more →

Google Public Sector boosts partner investments

🚀 Google Public Sector announced enhanced partner initiatives at Partner Connect to deepen collaboration and accelerate frontline adoption of mission-aligned solutions. Programs span enhanced incentives, a Partner-Led User Success Program, a Forward Deployed Engineer (FDE) academy, increased pre-sales deal funds, and $45M for post-sales deployment and adoption. Additional investments include an ISV Agent Factory, ISV ATO Accelerator funding, and new strategic alliances with SHI and Accenture Federal Services to expand market reach and operational readiness.
read more →

Google Cloud adds OpenAI models to Model Garden

🟢 Google Cloud announced that OpenAI’s newest models are now available in its Model Garden, expanding enterprise model choice on a unified platform. The integration lets customers select and swap models based on performance, latency, and cost, with usage billed through Google Cloud Marketplace and counted toward existing commitments. Google highlights governance and security via its Gemini platform and says select OpenAI GPT-6 variants are rolling out in private preview.
read more →

Google Cloud launches Confidential G4 with Blackwell GPUs

🔒 Google Cloud has made Confidential Computing on G4 instances with NVIDIA RTX PRO 6000 Blackwell Server GPUs generally available, integrating NVIDIA Confidential Computing into the AI Hypercomputer. These instances are now available across 36 zones on Compute Engine VMs and GKE nodes, with multiple consumption models and upcoming Confidential Space GA for secure multi-party data sharing. The offering provides hardware-enforced encryption, verifiable remote attestation, and near-native GPU throughput for agentic and inference workloads.
read more →

Claude Opus 5.5 and Sonnet 5.5 in Antigravity

🤖 Google Cloud now offers Claude Opus 5.5 and Claude Sonnet 5.5 in Antigravity for customers, enabling developers to pick models from the model picker after administrator enablement. The models integrate across Antigravity 2.0, CLI, and IDE extensions for VS Code, JetBrains, and Visual Studio, providing unified billing and security. The post outlines engineering patterns for multi-model workflows, including pairing implementer and judge models and using subagents for high-speed lookups and validation.
read more →

Gemini for Government Services previewed by Google Cloud

🛡️ Gemini for government services brings a governed, connector-driven AI agent to public sector workflows, integrating skills, secure system connectors, partner agents, and a central governance plane. Designed with agency leaders, it supports tasks like FOIA discovery, redaction, policy research, procurement review, and intake validation while preserving data sovereignty and compliance. The preview emphasizes role-bound access, auditability, and an open partner ecosystem for tailored deployments.
read more →

Gemini Enterprise Search on GDC Air-Gapped

🔒 This post announces enterprise search with Gemini on Google Distributed Cloud (GDC) air-gapped, offering a fully managed retrieval-augmented generation (RAG) search that runs entirely inside a sovereign perimeter. The preview supports multimodal uploads (documents, images, audio, video) and preserves conversations with source citations while keeping data and models local. Security features include AES-256 at rest with HSM keys, TLS with FIPS algorithms, local IAM/RBAC integration, and U.S. DoD IL6 alignment. The service embeds, indexes, and serves Gemini models on NVIDIA B300 GPUs within the customer's air-gapped network.
read more →

Google Distributed Cloud advances for sovereign AI

🔒 Google Distributed Cloud (GDC) air-gapped brings frontier and open-weight models, enterprise search with Gemini, next-generation IaaS, and an expanding ISV ecosystem directly into sovereign, disconnected environments. The update adds Gemini 3.5 Flash and support for models like Gemma and Mistral, A4 Ultra instances powered by NVIDIA Blackwell GPUs, enhanced VM networking, Cloud DNS, hardware-offloaded IPsec, and Elastic Security integration. Partners such as Teradata, Solo.io, and Mattermost provide data analytics, service mesh and secure collaboration to run natively within air-gapped deployments.
read more →

Python UDF logging enhances BigQuery pipeline observability

🛠️ This announcement introduces Python UDF logging in BigQuery, allowing developers to use Python's standard logging library inside UDFs to emit structured logs directly to Cloud Logging. The feature supports row-level diagnostic telemetry for fault-tolerant pipelines: malformed or corrupted rows can be quarantined (returned as NULL) while valid records proceed. Teams can create log-based metrics to trigger real-time alerts and query quarantined payloads via a linked _AllLogs view in BigQuery for triage and replay. The post walks through an e-commerce ingestion example and outlines best practices such as enabling debug output, avoiding sensitive data in logs, and sampling to manage costs.
read more →

Day‑0 support for Reflection Beam on Google Cloud

🚀 Google Cloud announces day‑0 support and optimized deployment recipes for Reflection’s new open‑weight model, Beam, enabling up to 5x higher throughput and 8x lower latency for agentic coding workloads. The post details integration with AI Hypercomputer, llm-d (vLLM), and NVIDIA Dynamo (SGLang) and describes techniques like prefix-aware routing, disaggregated prefill/decode, speculative decoding, and tiered KV cache offload to DRAM and local SSD. Recommendations, manifests, and benchmarking scripts are provided for running Beam’s BF16, FP8, and NVFP4 checkpoints on A4X Max and G4 instances to preserve model token efficiency at scale.
read more →

Architecting scalable agentic biotech workloads

🔬 This Google Cloud blog describes a three-layer reference architecture for building enterprise-grade, high-compute AI agents for drug discovery. It separates cognitive orchestration from GPU-bound execution and anchors both inside an enterprise governance plane. The post details execution choices (on-prem vs managed), containerized NVIDIA Inference Microservices on GKE, warm pools and preload strategies to avoid cold starts, and patterns for passing heavy payloads by reference. It also outlines Gemini platform governance features—agent identity, gateway, registry, and Model Armor—plus recommended asymmetric gating for reversible and irreversible actions.
read more →

Four practical evaluation engineering lessons

🔍 Evaluation engineering underpins reliable agentic systems and measuring behavior is harder than building the agent. Drawing on work developing the Google Cloud Developer plugin, the article presents four lessons: grade raw execution trajectories not final text, preserve turn structure in telemetry to catch parallel dispatch issues, run live and plan-only evaluations side-by-side, and treat evaluations as continuous integration against golden sets. The piece warns against one-shot demos and advocates stress-tested eval suites for safe deployment.
read more →

Fundamentals for Trustworthy Generative Media

🧭 Google Cloud Consulting outlines the four-part SAVE method to produce trustworthy, brand-aligned generative media. The guidance emphasizes separating fact extraction from creative generation, locking down core claims, and enforcing deterministic checks before and after generation. It advocates human-in-the-loop review at the script/storyboard stage and programmatic safety gates to reduce fabrications and preserve brand integrity.
read more →

Check Point integrates with Google Cloud Gemini Enterprise

🛡️ This post explains how Check Point extends Google Cloud’s Gemini Enterprise control plane to secure production AI agents through discovery, risk assessment, and runtime protection. It outlines integration points—Agent Registry, Agent Gateway, and observability—and describes how Check Point enforces prompt defense, MCP security, and context-aware policy enforcement. Examples from Dropbox and Nubank illustrate low-latency, high-accuracy protection at scale.
read more →

Borderless Lakehouse native Gemini integration

🔍 The article announces a preview integration between Gemini and Google Cloud’s borderless Lakehouse, enabling agents and users to query and reason over distributed enterprise data without moving it. It highlights new GA features — cross-cloud interconnect, Iceberg V3 support, read-write interoperability, AlloyDB federation, and partner integrations — to scale petabyte reasoning, reduce latency, and preserve security. The piece emphasizes improved metadata, caching, and zero-copy integrations to accelerate AI adoption.
read more →