Analysis
Benchmark partners said this week that the AI boom is straining the venture industry's long-standing informal rule against backing direct competitors, a rare public admission from a firm whose entire brand rests on near-exclusive founder loyalty. Speaking at TechCrunch Disrupt, general partner Eric Vishria told TechCrunch that "you are going to get conflicts, and that's just the reality of the current dynamic," while stressing Benchmark has avoided "that type of insane direct conflict" itself.
Pulse has previously covered Benchmark's bets in this cycle. The backdrop is concrete, not hypothetical. OpenAI and Anthropic now share dozens of overlapping investors, a level of cross-ownership that would have been unusual in prior software cycles. More pointedly, Khosla Ventures backs both Factory and Cognition, two AI coding-agent startups now competing directly -- Cognition makes the Devin agent. In late September, Factory fired board observer Chris Degnan, alleging "unethical conduct involving Cognition." Degnan denied sharing confidential information and has since become Cognition's chief revenue officer; Khosla founder Vinod Khosla sided with Degnan publicly, calling Factory "a struggling second-tier competitor."
“Pulse has previously covered Benchmark's bets in this cycle.”
## Why is AI accelerating this specific problem? Partner Chetan Puttagunta argued the dynamic is structural: as software costs "effectively go to zero," founders become "incredibly ambitious" and chase adjacent markets faster than in prior cycles, so category overlaps between portfolio companies are "probably going to happen more" rather than less. Peter Fenton called the issue "kryptonite" and the firm's toughest current problem, noting Benchmark sits closer to the co-founder end of the investor spectrum and treats transparency as a golden rule: "There shouldn't be the surprise after the fact."
That last line is the honest tell. Benchmark isn't claiming the conflict-of-interest norm is dead -- it's describing a firm trying to manage a problem it didn't fully see coming, by leaning on disclosure rather than avoidance. The firm has historically differentiated itself precisely by not doing what some multi-stage funds do routinely: writing checks into three or four companies chasing the same wedge and letting the market sort out which one wins.
However, nothing here has actually been tested in court or at a board, which is the real risk investors are glossing over. Vishria's comments are reassurance from a stage, not a policy change, and the Factory-Cognition episode shows how messy enforcement gets once a portfolio conflict turns into a personnel dispute rather than a polite LP conversation. For founders, the practical shift is diligence: a term sheet from a multi-stage fund now carries a real question about where else that fund's capital could end up competing within 18 months, not just where it sits today.
