OpenAI and Databricks sign $100 million deal to build enterprise AI agents

OpenAI and Databricks have signed a multiyear deal valued at $100 million to bring advanced artificial intelligence services to large business customers. The partnership aims to make it easier for enterprises to build AI agents that can leverage their own corporate data.

Under the agreement, OpenAI’s models—including its newest flagship, GPT-5—will be directly accessible to companies using Databricks’ data and analytics platform. This integration allows organizations to build AI agents on top of their existing data infrastructure without complex customization.

“It just works for them out of the box,” said Ali Ghodsi, co-founder and CEO of Databricks. “We think it’s going to drive significant incremental, actual revenue and usage.”

A Revenue Boost—and Bigger Ambitions

The two San Francisco-based companies expect the pact to generate at least $100 million in revenue over several years. But OpenAI Chief Operating Officer Brad Lightcap suggested the upside could be far larger.

The partnership also involves joint research and development efforts between OpenAI and Databricks, though neither firm disclosed detailed financial terms.

Lightcap called the deal OpenAI’s first true enterprise-focused data integration. “Data is the lifeblood of AI in the enterprise,” he said. “Our hope is that from experimentation to deployment, this partnership is able to accelerate enterprises, both with AI and data, as they look to build the next generation of software.”

Part of a Broader Push

OpenAI has been accelerating its business partnerships. This week, it unveiled plans for five new data center complexes in collaboration with Oracle and SoftBank, alongside a $300 billion cloud deal with Oracle. Nvidia also announced a $100 billion investment in OpenAI.

For Databricks, the OpenAI deal builds on existing collaborations with other AI model developers, including Anthropic. In March, Databricks announced a similar $100 million partnership with Anthropic to power enterprise AI agents.

Databricks, which reports serving more than 20,000 business customers, recently said it had surpassed a $4 billion annual revenue run rate.

Solving the AI Agent Challenge

AI agents—autonomous bots that can handle business tasks—are still in the early stages of enterprise adoption. Reliability and data integration remain hurdles, with many companies hesitant to deploy AI tools that cannot deliver high accuracy.

Ghodsi said Databricks’ research teams are working to push agent accuracy levels above 95%, which he considers on par with human performance.

Mastercard is already experimenting with Databricks-powered AI agents. Greg Ulrich, the company’s chief AI and data officer, said Mastercard uses them in areas such as customer onboarding and support. “The key to getting value out of AI is to be able to link it with your data,” Ulrich noted.

The Bigger Picture

Software vendors from Salesforce to Workday are also racing to combine enterprise data access with AI agent technology. OpenAI, while offering its own business-facing platform, lacks direct data pipelines—making Databricks an ideal partner.

For enterprises, the partnership signals a maturing AI ecosystem where data integration is becoming just as critical as the AI models themselves.