Startup Slack Threads Become Commodity for AI Training
AI training companies like Mercor are actively acquiring internal communications and code from startups, offering payments up to $300,000 for Slack threads, GitHub records, and meeting transcripts. Warmly's CEO received four such acquisition offers within days of the company's HubSpot acquisition announcement. The practice highlights how internal startup data has become a commodity for AI model training, even as acquirers may not want the same datasets.
TL;DR
- Mercor and similar AI training firms are buying or licensing startup internal communications, code, and meeting records for AI model training
- Warmly received four unsolicited offers for its datasets after announcing its HubSpot acquisition in late June
- Companies are specifically targeting data that traditional acquirers exclude from deals, including Slack messages, GitHub content, Asana records, and staff meeting transcripts
- The practice reveals a new market for startup operational data driven by demand from AI labs including OpenAI, Anthropic, and Google
Why It Matters
This signals a structural shift in how AI training data is sourced and valued. Startups now face a secondary market for their internal communications, creating both a revenue opportunity and a privacy consideration. The fact that AI labs are willing to pay for this data suggests it has meaningful training value, even if it comes from failed or acquired companies.
Business Impact
For startup founders and investors, internal communications now represent a monetizable asset separate from the company's core business value. However, this also introduces new legal and contractual complexity around employee privacy, confidentiality agreements, and what data can be sold post-acquisition. Companies need to understand their obligations before accepting such offers.
Key Implications
- Startup data, particularly internal communications and process documentation, is becoming a distinct asset class with market value for AI training
- Employees and contractors may not expect their internal communications to be sold to third parties, creating potential privacy and trust issues
- Acquirers may need to explicitly negotiate for or against the inclusion of internal datasets in acquisition agreements to avoid post-deal complications
What to Watch
Monitor whether this practice becomes standard in acquisition negotiations and whether regulatory scrutiny emerges around employee data privacy in these transactions. Track whether startups begin monetizing this data as a revenue stream and how employee agreements evolve to address data ownership and consent.
Subscribe to the newsletter
The latest stories and analysis, delivered to your inbox.
Free. No spam. Unsubscribe any time.
