Google Cloud announces new AI processors and enterprise agent platform, shifting its AI chip strategy

Google disclosed changes to its AI processor (the custom-designed chips that run AI models) strategy at its cloud computing event. The company is introducing new hardware accelerators alongside an enterprise agentic platform—a set of tools that let businesses deploy AI agents (AI systems that take actions on their own) to automate workflows without constant human direction.

The new accelerators are built to handle inference (the step where an AI produces an answer) more efficiently, meaning businesses running AI on Google Cloud will process requests faster and at lower cost per query. The enterprise agentic platform lets non-technical teams define business processes once, then let AI agents execute them repeatedly—reducing manual setup time.

For business professionals using Google Cloud, this matters because it lowers the barrier to deploying AI automation. Teams no longer need machine-learning expertise to build AI workflows; marketing, operations, and finance teams can now set up AI agents to handle routine tasks like data processing, customer outreach, or compliance checks. For enterprises currently locked into manual processes or expensive consulting-led AI projects, cheaper processors and simpler tools mean faster ROI and faster time-to-value.