Currently, the most effective alignment happens when organizations solve for access and control,” says Arun Sundaram, director of Google services and solutions at CDW. By already launching new support for the Wiz platform, Google appears to be ready for some major new security platform launches potentially later this year. As part of Google Cloud now, Wiz also launched a new security technology for securing AI-native development, including new protections for vibe-coded applications via an integration with Lovable. The TPU 8t is optimized for training, designed for high-throughput AI workloads, while Google’s new TPU 8i is optimized for inference and reinforcement learning. The platform brought model selection, model building and agent building capabilities together with new features for agent integration, DevOps, orchestration and security. On the innovation front, Google Cloud has been focused on building new or expanded agentic and AI agent solutions in 2026.
Throughout deSouza’s nearly four-decade IT career, he’s held top positions including president of product and services for Symantec and CEO of biotechnology firm Illumina. The company announced last month a $9 billion investment in South Carolina to expand its Berkeley County data center campus and support the continued construction of two new sites in the region. In October, Google unveiled a $15 billion investment in India to build a new AI-powered data center as part of a massive undertaking to create a global network of Google AI data centers. Google has been one of the largest spenders on building new data centers to power its cloud and AI technology as well as increase its customer reach. Google Cloud launched dozens of new cybersecurity integrated solutions on the marketplace this year from security leaders like CrowdStrike, Palo Alto Networks and Menlo Security. For AI agent builders, the marketplace now offers co-selling opportunities and better global customer reach to help builders monetize AI innovations through Google Cloud’s global distribution.
And canvas allows users to create and edit files without switching apps, according to Google. Users can deliver these agents through the Gemini Enterprise application, which has been updated with a new agent designer feature for schedule- or trigger-based agents, long-running agents for complex processes and an inbox for managing agent activity. During Next, Google also revealed support for the latest Gemini 3 models in air-gapped and connected environments, https://innovatenexes.com/crafting-efficiency-with-mobile-devices.html which could appeal to solution provider customers in financial services, health care and other regulated industries. Agent Platform gives users access to Google’s most advanced AI model, Gemini 3.1 Pro, plus other Google first-party models and models from Anthropic, including Claude Opus 4.7. Agent anomaly detection flags suspicious agent behavior, tools misuse, unauthorized data access and reasoning drift.
- Effective in January, Google Cloud no longer charges customers’ egress fees for moving data from its infrastructure to another cloud provider or migrating workloads to on-premises data centers.
- In fact, 70% of enterprise users who use “Help me write” in Docs or Gmail end up using Gemini’s suggestions, and more than 75% of users who create images in Slides are inserting them into their presentations.
- As for TPU 8i, there’s now support for up to 1,152 TPUs per pod via Boardfly Topology.
- They launched new tools that let anyone build these AI helpers without needing to know how to code.
- Some of the biggest AI product launches this year from Google Cloud were around Gemini innovation and new TPUs tensor processing units.
Expanding Vertex to Enable a Multi-Agent Ecosystem
- This will be made available via a new conversational analytics API, now in preview, so data teams can embed this easy question and answer layer into their existing applications.
- They’re launching new, powerful computer chips to handle all this extra work and keep systems secure from hackers.
- We’re moving onto Gemini Code Assist, Google Cloud’s AI pair programmer, which Calder says is already being used by a wide range of enterprises.
- Google Threat Intelligence is seeing upgrades today, with new agentic capabilities that shift the burden of data synthesis and initial artifact triage to a specialized suite of AI agents built with the newest Gemini models.
- Papa John’s is using our Ordering Agent, which uses AI context to remember “the usual” order and get diners what they want faster.
The agent features a unified voice-and-text AI ordering system, , while integration with the Omnichannel Gateway helps the agent remember customer preferences and past orders and get food to the door faster. Citadel Securities built a scalable, cloud-based research environment that uses TPUs to run AI workloads up to 4x faster with 30% lower costs. As the evolution of Vertex AI, it offers the model selection, model building, and agent building capabilities that customers love, plus new features for agent integration, DevOps, orchestration, and security. Google’s plan is to improve Google’s Cloud cybersecurity technologies and capabilities around multi-cloud systems. The new Gemini Enterprise securely connects to customers’ data wherever it lives, including Google Workspace, Microsoft 365 and Microsoft SharePoint to business applications like Salesforce and SAP. Gemini Enterprise brings together several AI technologies including Google’s Gemini models, first- and third-party AI agents, and core conversational AI and orchestration innovation from Google technology formerly known as Agentspace.
Google Cloud unveils agentic AI security strategy with Wiz integration and threat intelligence upgrades
Google Cloud is a leader in cloud infrastructure, with deep AI expertise and a track record of industry-leading security innovation. Organizations of all sizes — from start-ups and large enterprises to governments and public sector organizations — can use Wiz to protect everything they build and run in the cloud. Wiz delivers an easy-to-use security platform that connects to all major clouds and code environments to help prevent cybersecurity incidents. Both cybersecurity and cloud computing are rapidly growing industries with a vast range of solutions. Dan has covered technology for three decades and held senior editorial positions with Ziff Davis, Jupitermedia, Disney Publishing, McGraw-Hill and Advance Digital.
We engineered Virgo Network with autonomous reliability capabilities to maximize workload efficiency at scale, a.k.a., goodput. A single stalled instance can stop an entire synchronous training job, wasting valuable compute cycles. Virgo Network delivers up to 4x the bandwidth per TPU 8t accelerator over the previous generation, and 40% lower unloaded fabric latency for TPU 8t compared to the previous generation network for TPUs.
The Mountain View, Calif.-based cloud giant has been growing sales at a fast pace in 2025 while investing billions in AI infrastructure and launching a slew of new AI product innovations—from Agent2Agent Protocol to Gemini Enterprise. From Google’s $32 billion upcoming acquisition of Wiz to rapid Gemini and agentic AI innovation, there was no shortage of huge news stories from Google Cloud throughout 2025. Just this month, federal judge Amit Mehta delivered a nuanced ruling in the government’s five-year-old search monopoly case, attempting to curb Google’s dominance without hampering its AI ambitions.
And Google’s Agentic Taskforce now has Workspace Intelligence for streamlining productivity through a unified context layer across Workspace apps, among other new capabilities, according to the vendor. The lakehouse also enables zero-copy access to Databricks, Salesforce, ServiceNow, Snowflake and other third-party apps, operating systems and platforms. And Google has expanded its compute services frameworks and inference engines to natively support Pytorch for TPUs. Virgo Network for A5X supports up to 80,000 GPUs in a single data center and up to 960,000 GPUs across multiple sites. Virgo Network and TPU 8t can connect 134,000 TPUs into a single fabric in https://flarealestates.com/what-you-need-to-know-about-modern-technologies-and-artificial-intelligence-in-trading.html a single data center and connect more than 1 million TPUs across multiple data center sites into a training cluster. In networking, Google’s AI-optimized Virgo can connect Nvidia Vera Rubin NVL72 systems or TPU 8t superpods into supercomputers with hundreds of thousands of accelerators, according to Google.
These advanced AI assistants work to automate tasks autonomously, as Kurian explains. This can be combined in the platform to create quick clips for advertising and more. We’re told his mission is to create a trailer for the party to end the event – complete with a gag about Kurian not wanting to be able to sing Chappel https://www.internetling.com/2019/12 Roan but not getting permission. In the video, customers say that Gemini is already cutting down their toil and opening new time for valuable work.
For example, customers like Bayer, Mayo Clinic, Mercado Libre, NewsCorp, and Vodafone are already seeing benefits. Now, data teams can use Gemini models for multimodal and advanced reasoning for their existing data. Google Cloud lets you combine the best of AI with your grounded enterprise data, while keeping your data private and secure. We are proud to share that Gemini Code Assist supports your private code base to be anywhere — on premises, GitHub, GitLab, Bitbucket, or even multiple locations. We deployed it to a group of developers inside Google and found they had more than 40% faster completion time for common dev tasks and spent roughly 55% less time writing new code.