Y Combinator founder Paul Graham says AI-written founder emails feel like being lied to

The more pointed issue isn't that AI emails feel inauthentic, it's that Graham is effectively setting a social norm for YC's intake funnel, one that disadvantages founders who lean on LLM assistance for outreach regardless of whether their underlying idea is strong.

This is largely disconnected from recent activity in our archive, as we have no prior coverage to anchor it to. That said, it belongs to a growing pattern visible across the broader industry: the gap between what LLMs can produce and what high-trust professional contexts will accept. The friction Graham describes is not about output quality but about signal integrity, the same dynamic playing out in recruiting, legal correspondence, and board communications where authenticity functions as a proxy for judgment. His dual role as YC gatekeeper and early OpenAI backer makes the tension unusually legible: he helped build the tools he is now warning founders not to misuse in contexts where voice and intent are the actual product being evaluated.

Watch whether YC formalizes any guidance to applicants about AI-assisted correspondence in the next application cycle. If it does, other top-tier accelerators and seed funds will likely follow within two cycles, hardening an informal norm into explicit selection criteria.