A DOJ probe of a16z board seats asks whether venture portfolios are interlocking directorates

A board seat is a negotiated right in a venture deal, priced into the terms and defended round after round. It is also the part of the arrangement a firm cannot quietly unwind once the market moves, which is where this case gets awkward. The seats at issue, as described on TechCrunch's Equity podcast, are Ben Horowitz at Databricks and Martin Casado at Fivetran, two companies that now compete with each other ^[2]. Nobody bought into a rivalry: TechCrunch notes the companies were not necessarily direct competitors when a16z first invested ^[5], and Kirsten Korosec pointed out that in this AI cycle plenty of companies have changed what they do, with a firm holding as many seats as a16z likely to be sitting on both sides when they do ^[6].

So the trigger, if the DOJ has one, is a portfolio company's roadmap. Diligence at entry does not catch that, and no investor controls it afterward. What a fund does control is how many seats it takes inside one category and whether it writes the second check at all. The statute being used is old and lightly exercised: TechCrunch describes a 112-year-old antitrust law rarely applied to VCs ^[9], which places its enactment around 1914 ^[19]. Anthony Ha, who spent a couple of years at a much smaller early-stage firm ^[18], said the rule against sitting on competitors' boards exists but is not closely enforced ^[11]. Founders have policed it socially instead, since nobody enjoys watching their investor take a seat with their biggest competitor ^[16].

The politics do not settle the read. O'Kane noted the inquiry began under the Trump administration ^[4] and has run close to a year ^[3], while a16z's leadership is friendly with that administration and sits on some of its councils ^[13]. The same administration settled with Live Nation without breaking up Ticketmaster, which O'Kane took as evidence that its antitrust rhetoric has not borne out ^[14]. Korosec's question is the one the firm's peers are asking: of everything the DOJ could put at the top of the pile, why this ^[15]. A year in, the outcome is not known ^[17], and a16z has said little, in contrast with the Biden years, when by O'Kane's account every small policy change generated a day of posting ^[12].

The deterrent operates before any filing. O'Kane's speculation is that the DOJ is making an example that smaller firms will follow ^[20], and that is the mechanism most likely to bite. A firm with counsel on staff can absorb a year of quiet scrutiny; a seed fund with three positions in the same tooling category hears about this from its own lawyers long before it hears from Washington, and the cheapest response is to give up the seat or pass on the second deal. That is a real change in how a category gets financed, produced by a theory nobody has seen stated. TechCrunch framed the operating question plainly: how do you manage board seats when the boundaries between your portfolio companies keep moving ^[8]. Until the DOJ answers or closes, the safest assumption in a partnership meeting is that the boundary will move, and that the seat will be the thing regulators look at.