a16z Partners Play Both Sides While DOJ Watches Closely

Andreessen Horowitz, the self-appointed intellectual godfather of Silicon Valley, has discovered an innovative new business model: letting your partners sit on the boards of companies that compete with each other, then acting surprised when federal regulators notice. Ben Horowitz chairs Databricks while Martin Casado does the same at Fivetran—two data infrastructure firms locked in direct competition—a fact the Department of Justice has apparently found interesting enough to investigate for nearly a year. The arrangement is so transparently problematic that one wonders if a16z's legal team simply didn't read past the first paragraph of the conflict-of-interest manual before filing it away.

For those unfamiliar with the data engineering arms race, Databricks and Fivetran both operate in the increasingly crowded space of data pipelines and ETL (extract, transform, load) infrastructure—the plumbing that moves data around corporate networks like blood through a venture capitalist's wallet. These aren't niche players with complementary features; they're direct competitors pursuing the same customers, the same use cases, and the same pools of venture capital. Placing senior a16z partners on both boards creates a straightforward incentive structure: whatever information Horowitz learns at Databricks could theoretically inform a16z's strategy at Fivetran, and vice versa. The DOJ's nearly year-long investigation suggests someone at Justice felt this warranted a closer look than a standard Saturday morning review.

This is hardly a16z's first rodeo with the kind of structural conflicts that make compliance officers nervous. The firm has built its reputation on aggressive portfolio concentration—backing multiple companies in the same market segment, then claiming synergies justify the overlap. Their partners sit on numerous boards simultaneously, creating a web of interconnected interests that would make a spider's head spin. When you're managing a $28.3 billion fund with partners embedding themselves across dozens of board seats, the assumption appears to be that good intentions and internal governance will prevent any mishaps. History suggests otherwise.

The official a16z position on such matters typically sounds something like this: we believe strongly in operational excellence and maintaining the highest ethical standards while simultaneously pursuing opportunities to create defensible competitive advantages through strategic portfolio alignment. Translation: we think we can manage this, and besides, everyone else is doing it. This is the venture capital version of 'trust us, we know what we're doing'—a phrase uttered by countless executives in the moments before regulatory agencies arrive with subpoenas.

The real risk here isn't just legal exposure, though a16z can probably afford whatever settlement the DOJ eventually negotiates. The risk is reputational: a government investigation into conflicts of interest at the industry's most prominent megafund sends a signal to LPs, founders, and regulators that even the untouchables aren't above scrutiny. If the DOJ determines that a16z partners used their board positions to coordinate strategy between competing portfolio companies, the implications ripple far beyond Databricks and Fivetran. It suggests that venture capital's entire honor-system approach to conflict management deserves federal attention.

This moment represents a broader reckoning: as VC firms have grown larger and consolidated more power, their ability to police themselves has shrunk in inverse proportion. The assumption that market forces and partnership agreements can adequately manage conflicts of interest looks increasingly quaint. What a16z is learning the hard way is that the 'we'll figure it out later' approach to governance works fine until the Department of Justice decides the industry's self-regulation deserves a closer examination.