Micro1 Is Spending $1 Billion to Buy the Workplace

Micro1 is making a $1 billion bet that the next valuable AI training material is not sitting on the open web but inside ordinary companies’ archives. Business Insider reports that the startup plans to spend that money over the next year acquiring and licensing internal business data for AI model development, with payouts of $100,000 to $2 million for a single midsize company’s dataset.

That is a startling price tag for what many firms would have treated, until recently, as digital sediment: Slack messages, CRM logs, wiki pages, design files, checklists, the daily residue of people trying to get through work. Micro1’s pitch is that this residue is useful precisely because it is connected. William Almeida, the company’s chief revenue officer, told Business Insider that Micro1 usually wants a company’s entire data corpus so an AI model can replicate what it would be like to be a human employee. The word “entire” is doing a lot of work there.

Business Insider says Micro1 works in three ways. It can take raw company data, transform it to protect privacy, and license it to an AI lab. It can provide an in-progress AI agent to a company and collect feedback on how it performs. Or it can build a simulation by taking a company’s data up to a certain timestamp, freezing that environment, and giving it to a model developer trying to teach AI how to work from existing context.

In less polished language, the company is trying to turn office history into training terrain.

Ilya Brotzky, the CEO and cofounder of recruiting startup VanHack, told Business Insider he was shocked when Micro1 made what he called a “substantial” offer after he filled out a survey about his company’s data. “I didn’t believe it, to be honest,” he said. “I felt like it was a little too good to be true.” After looking into Micro1’s privacy process, Brotzky said he felt comfortable sharing the requested material, including VanHack’s Google suite, Confluence data, Slack messages, and customer relationship management software.

Grant Gooley, the CEO of MyPortal Marketing, told Business Insider he sold internal design files from Figma and Canva, along with checklists from Monday.com. “When we were building these processes years ago, I didn’t think that we would be using it to support AI learning,” he said. “But doing that, it’s very fulfilling.”

The broader market logic is easy to see. Business Insider places Micro1 in a new class of data brokers feeding AI labs such as OpenAI and Anthropic as those companies look beyond scraped internet text. Gartner, as cited by Business Insider, estimated that spending on AI data is likely to rise from less than $1 billion last year to almost $6.5 billion in 2027. Micro1 is not dabbling around the edges of that trend. It is trying to industrialize it.

What makes this moment uneasy is not that the data sounds glamorous. It mostly does not. What matters is that mundane workplace systems capture process: how tasks move, how decisions get documented, where delays happen, what gets corrected, what gets missed. Almeida told Business Insider that the connection, diversity, and traceability of those data points matter for model training. That sounds right. A pile of isolated documents is one thing. An organization’s operating memory is another.

Business Insider also notes the people least involved in these transactions may be the ones who generated much of the material. Current and former employees are not likely to get compensation from, or much say in, their work becoming AI training data, the report says. Ifeoma Ajunwa, a law professor at Emory University, previously told Business Insider that such data sales are an “ethically gray area.”

That phrase lands harder here than the usual AI hand-wringing. These are not public posts volunteered into the chaos of the internet. They are internal traces, created for one purpose and now being repurposed for another, with a startup racing to convince executives that hesitation means leaving money on the table. Almeida told Business Insider he pitches the business as being in its early days: “If you wait too much, you won’t get the premium that the first companies are getting.”

A billion dollars can make a lot of archived office chatter look newly precious.