When a $29 billion company tells you it's automated away most of its support department, that's worth paying attention to. Cursor, the AI coding startup, has done exactly that—and its CEO is happy to explain how.

The Help Desk That Helps Itself

Speaking at Fortune's Brainstorm AI conference in San Francisco, CEO Michael Truell revealed that Cursor has built an internal AI help desk that resolves the majority of employee and customer support requests without any human touching them. According to Fortune, the results are dramatic.

"We've already automated roughly 80% of our support tickets," Truell said. The system isn't just a generic chatbot either—it's been customized to work with Cursor's internal knowledge base and operational tools, which presumably makes it far more useful than the "press 1 for more options" systems we all love to hate.

Ask the Company Brain Anything

Cursor didn't stop at support tickets. Truell said the company has also deployed an AI-powered internal communications system that lets employees ask questions and get immediate answers pulled from across the entire organization.

"We have a system where folks can ask any question about the company and get it answered by an AI," he explained.

To push automation even further, Cursor has embedded "forward-deployed engineers" who build custom AI tools for operations and sales teams as the company experiments with where else automation might work.

Does AI Actually Make Developers Faster?

Here's where things get interesting. A study by the nonprofit METR found that experienced developers sometimes took longer to complete tasks when using AI, partly because of time spent crafting prompts and reviewing the AI-generated code.

But a University of Chicago study found the opposite—teams using Cursor tools merged more code changes than non-users.

"A lot of folks think that junior developers get the most out of AI," Truell said. "It looked like senior engineers actually were more effective."

The Workplace Disruption Everyone's Talking About

Anthropic's internal research on Claude Code found similar complexity: productivity went up, but so did concerns. Employees completed more work and launched new projects, but many reported collaborating less with colleagues and worrying their technical skills were deteriorating. Job security anxiety increased too.

OpenAI CEO Sam Altman weighed in on the broader picture, saying AI will likely automate 30% to 40% of work tasks in the near future. He argued the technology would reshape how jobs function rather than simply eliminate them, while stressing the need for stronger regulation and safety measures as AI capabilities advance.

Meanwhile, Kevin O'Leary pushed back against fears of mass job losses, saying AI was improving work by removing repetitive tasks. He said many of his companies had adopted AI to cut costs and boost productivity, comparing the transition to earlier technology shifts like television and radio.

So is this the future of work? If Cursor's experience is any indication, we're about to find out whether "80% automated" becomes the new normal—or just another Silicon Valley experiment that doesn't scale beyond tech companies.