Databricks launches always-on AI agents after Anthropic office push

By Grace Kim

One day after Anthropic released corporate productivity tools, Databricks launched always-on AI agents—software that performs tasks autonomously—for enterprises on Wednesday, intensifying the contest to automate knowledge work and expand the pool of paying business customers.

What did Databricks put into work?

Databricks introduced two components on Wednesday, aimed at continuously analyzing and organizing the reams of data scattered throughout a business. The company says the capability will transform how office workers do their jobs.

• Genie assistant: The updated assistant runs in the background, follows a worker throughout the day, makes proactive suggestions and retrieves information based on current activity.
• Ontology: The separate feature scans organizational data, including management reporting lines and projects across departments, so employees can quickly access information or run analyses.

“The agents are running in the background all the time. So it’s always on,” said Databricks CEO and cofounder Ali Ghodsi. “I personally use it every day. Every hour of the day. It’s basically made me a 10x better CEO than I was a year ago.”

Ghodsi expects both large and small companies to use the products, extending Databricks’ target beyond technical teams.

Takeaway: Databricks is making persistent assistance the central product promise for enterprise software.

What is Anthropic’s enterprise AI push worth?\n

More than $40 trillion in global knowledge-worker value in 2025 anchors Anthropic’s enterprise AI pitch in the prospectus for an IPO (initial public offering) expected next week.

“We believe the work that knowledge workers perform represented more than $40 trillion in economic value globally in 2025, and with frontier AI’s potential to make them meaningfully more productive, our near-term opportunity is to expand this value and accelerate economic growth,” Anthropic said.

The day before Databricks’ launch, Anthropic, the maker of the Claude models, released productivity tools for corporate users:

• An AI-powered spreadsheet application competing with Microsoft Excel and Google Sheets.
• A feature that lets Claude join video calls, listen to the conversation and begin work while participants are still talking.

The IPO could value Anthropic at $1.4 trillion. A strong debut on Wall Street could serve as a positive signal for the broader AI industry. By comparison, privately held Databricks raised $5 billion at a $190 billion valuation in August.

AI vendors are trying to make deeper inroads into corporations and broaden their base of paying customers. The technology has already transformed software programming, where it has become an indispensable tool. In other business functions, adoption remains experimental and return on investment is less clear-cut. Autonomous agents and advances in AI model capabilities underpin industry expectations that the market can now reach a broader group of workers.

Takeaway: The near-term test is whether always-on agents can produce clear returns in business functions where AI remains experimental.

What risk remains around Nvidia?

The enterprise-agent push arrives with a separate leadership risk at Nvidia. The AI industry has been shaken by the revelation of CEO Jensen Huang’s extramarital affair with a woman whom U.S. officials believe is a spy for China.

Nvidia’s chips and financing have broad importance across the AI industry, raising fears that a forced departure by Huang could create a ripple effect. Ghodsi brushed off that scenario, describing Huang as an amazing CEO who should remain in place after building Nvidia for the last many decades. He expects the industry to weather the storm because the AI revolution is too large to be affected by the affair revelations.

Takeaway: A potential Huang departure could send ripples through chip supply and AI financing, the infrastructure supporting the wider enterprise-agent market.