Google Cloud chief executive Thomas Kurian says enterprises are betting on AI agents rather than AI applications, arguing that autonomous software agents will define the next phase of workplace technology.

In an interview this week, Kurian said companies are moving beyond chatbots toward agents that can execute multi-step tasks across business systems, from processing invoices to resolving customer tickets. He said Google Cloud’s enterprise customers are deploying Gemini-powered agents that connect to internal databases and third-party tools through open protocols.

Kurian pointed to agent-to-agent communication standards as a critical building block, allowing agents built by different vendors to coordinate on complex workflows. Google has been pushing interoperability frameworks alongside rivals including Microsoft and Salesforce.

The comments reflect a broader industry shift. Enterprise software spending on AI agents has grown sharply this year as firms seek measurable productivity returns from their AI investments. Analysts caution, however, that security, access controls and reliability remain hurdles before agents handle sensitive operations unsupervised.

Google Cloud has positioned its agent development platform as central to its competition with Amazon Web Services and Microsoft Azure for enterprise AI workloads.