Enterprises are entering a new phase of AI adoption, moving beyond pilots and chatbots towards agentic systems that can reason, plan and take action across real business workflows.

Databricks’ 2026 State of AI Agents report draws on fully aggregated, anonymised telemetry from more than 20,000 organisations worldwide – including over 60% of the Fortune 500 – to examine how AI agents are being deployed in production environments.

The findings show a sharp acceleration in multi-agent use, alongside a growing focus on governance, evaluation and enterprise-grade infrastructure.

Mult-agent system: The next frontier?

Enterprises are transitioning from single chatbots to multi-agent systems that autonomously orchestrate end-to-end workflows.

Databricks found that usage of multi-agent workflows on their platform has grown by 327% in just four months.

Technology companies, especially digital natives, are building multi-agent systems nearly four times more than any other industry, reflecting early enterprise maturity.

AI is critical to workflows

Enterprises are taking a pragmatic approach to AI by solving real-world business problems specific to their industry. Manufacturing and automotive industries typically use AI for predictive maintenance (35%), retail and consumer goods opt for market intelligence (14%), while health and life sciences use AI mostly for medical literature synthesis (23%).

Four out of ten of the top AI use cases focus on practical customer concerns such as customer support, advocacy, and onboarding.

Businesses are increasingly using multiple LLM model families (such as ChatGPT, Claude, LLama, and Gemini), aligning their use cases to those that perform best for a given task and maintaining vendor flexibility. As of October 2025, 77% of customers were using two or more, while the share of those using three or more rose from 36% in August to 59% in October.

Evaluation and governance

Databricks found two major factors in common for companies moving from generative AI experimentation to getting it into the real world.

Businesses that actively use AI governance put 12x more AI projects into production. Having  unified governance dictates how data is used, setting guardrails and establishing accountability.

Customers that use evaluation tools put 6x more AI projects into production. These are frameworks that measure, test and improve the quality and reliability of AI models at all stages of deployment.

Agents in database operations

AI agents now create 80% of databases, up from near zero just two years ago.

Nearly all (97%) of database testing and development environments are now built by AI agents. Agents are dramatically reducing the time needed to clone, branch, and test databases, supporting faster experimentation and deployment.

With the rise of “vibe coding” business users without deep technical expertise can create AI apps and help democratise the technology across the company.

Over 50,000 data and AI apps have been created since the Public Preview of Databricks Apps, with a 250% growth rate over the past six months.

Recommended reading

• Will AI Agents Be the Enterprise Disruptor of 2026?
• Can AI Agents Replace White-Collar Workers?
• Microsoft Tests AI Agents in Simulated Marketplace
• 80% of Firms Say Their AI Agents Have Taken Rogue Actions

“For businesses across EMEA, the conversation has moved on from AI experimentation to operational reality,” Dael Williamson, EMEA CTO at Databricks, said.

“AI agents are already running critical parts of enterprise infrastructure, but the organisations seeing real value are those treating governance and evaluation as foundations, not afterthoughts.

“Just as importantly, competitive advantage is shifting back towards how companies build – not simply what they buy.

“Open, interoperable platforms allow organisations to apply AI to their own enterprise data, rather than relying on embedded AI features that deliver short-term productivity but not long-term differentiation.

“In highly regulated and risk-aware markets across EMEA, that combination of openness and control is what separates pilots from competitive advantage.”