Google to Spend Up to $185 Billion on AI and Cloud

Google plans to invest between $175 billion and $185 billion in capital expenditures this year to expand data centers, buy AI chips and build the hardware and software needed to run large models and ‘agentic’ workflows, CEO Sundar Pichai announced at Google Cloud Next in Las Vegas. The figure contrasts with roughly $31 billion in capital spending in 2022.

The company described ‘agentic’ systems as AI agents that can carry out complex tasks with limited human oversight. Pichai said Google is already using those systems in production and cited an increase in AI-assisted software development: nearly 75% of new code at Google is now generated by AI and approved by engineers, up from about 50% last fall.

Google said the budget will fund new data centers, custom AI processors and other infrastructure to train and run frontier models and production agents. The company presented the spending as part of its cloud strategy and highlighted commercial partnerships and customer deployments it expects will use that infrastructure.

To help partners build and sell agentic products, Google announced a $750 million fund for its 120,000-member Google Cloud partner network. The initiative will offer engineering support, early access to Google’s Gemini models and financial incentives for major consulting firms including Accenture, Deloitte and McKinsey.

Financial services firm Citi introduced ‘Citi Sky,’ an AI-powered wealth management assistant built on Google’s infrastructure. AI developer Thinking Machines Lab said it has expanded use of Google Cloud’s AI Hypercomputer to accelerate model training.

Google also described internal uses of AI beyond code generation. Company teams use automated tools to triage unstructured threat intelligence, extracting key details and filtering noise. Pichai noted that automation has reduced threat mitigation time by more than 90% in their security operations center.

Executives framed the spending as a way to make Google Cloud the underlying infrastructure for AI developers and enterprise customers, offering hardware and models to train and run advanced systems. The company emphasized partner incentives and customer deployments as pathways to commercial returns on the infrastructure investment.

The announcements come amid large investments from other cloud and AI providers. Google used the conference to describe a shift from chat-style systems to agents that act on behalf of users and to highlight the operational challenges of deploying and managing many agents at scale.

Quote: Pichai said, ‘Nearly 75% of all new code at Google is AI-generated and approved by engineers, up from 50% last fall.’