B Capital is betting that AI-augmented investing is a product line, not just an internal efficiency play. The firm hired Andrew Jackson, formerly of G42, as its first chief AI officer, with a mandate that spans deal diligence, tool development, and eventually spinning out its Bee Hive investment platform, the AI system that already drafts most of B Capital’s investment committee memos, to other firms. A spinout would strip out proprietary data and partner context, selling only the underlying model and tooling. The hire follows similar moves at Eclipse, Felicis, and Thrive Capital, and lands as B Capital closes a $2.8 billion acquisition of Russell Investments with AI expected to play a central role in that integration<\/a>.<\/p>\n The story that matters here isn’t the hire itself, it’s the spinout intent. A VC firm building AI tooling for its own operations and then packaging it for external sale is a familiar pattern in enterprise software history: internal necessity becomes a product. If B Capital actually ships Bee Hive externally, asset managers, family offices, and corporate development teams running M&A pipelines would face a purpose-built AI diligence layer built on real deal data, not a generic LLM wrapper. Whether your firm is a potential customer or a potential competitor to that product depends on how much of your own deal or investment workflow is still running on spreadsheets and email threads.<\/p>\n Jackson’s most interesting observation isn’t about AI autonomy, it’s about the asymmetry in how investment decisions get recorded. His framing, that roughly 90 percent of investment decisions are passes and almost none of that data gets captured, points to a broader failure mode in knowledge-intensive enterprise functions. Legal teams lose the reasoning behind contract positions that weren’t taken. Procurement teams rarely log why a vendor was cut from a shortlist. HR teams seldom record why a candidate was passed at the resume stage. Every function that makes high-volume, low-documentation decisions is sitting on the same structural gap B Capital is trying to close, and the firms that build systems to capture the “no” decisions will have a compounding data advantage over those that only track the “yes.”<\/p>\n B Capital’s co-CEO framing of current AI capability as “a very smart summer intern” is useful calibration, not modesty. It maps directly to the governance question CIOs are fielding right now: at what confidence threshold does AI move from drafting and flagging to deciding? B Capital is drawing the line at committee memos and note-taking today, with a stated aspiration to give AI a non-voting seat eventually. The firms that will regret that progression are the ones that let the line drift without a deliberate decision to move it. The leading indicator to watch isn’t when AI gets a seat at the table; it’s whether your organization has a documented policy for when it doesn’t.<\/p>\nWhat this means for your business<\/h2>\n