Working Paper 26-090

AI-Native Firms

Hyunjin Kim
Rembrand Koning

AI-Native Firms
Hyunjin Kim
INSEAD

Rembrand Koning
Harvard Business School

Working Paper 26-090

Copyright © 2026 by Hyunjin Kim and Rembrand Koning.
Working papers are in draft form. This working paper is distributed for purposes of comment and discussion only. It may
not be reproduced without permission of the copyright holder. Copies of working papers are available from the author.
Funding for this research was provided in part by Harvard Business School.

AI-Native Firms∗
Hyunjin Kim
INSEAD

Rembrand Koning
Harvard Business School
June 9, 2026

We study how firms built around AI capabilities—“AI-native” firms—are organized. Drawing
on Y Combinator batches W20–F24 and U.S. venture-backed startups whose first financing
closed between 2020 and 2024, we classify each firm’s AI-native status and link it to workforce
microdata on team size, function, seniority, and hierarchy. Relative to non-AI startups in the
same industry-cohort, AI-native firms are 25% smaller. Their share of engineers is 13% greater,
and the shares of entry-level workers and managers are each roughly 15% lower. Their hierarchies
are half a seniority level flatter—yet valuations are comparable, suggesting higher value created
per employee. We argue these patterns reflect two channels: a process channel, in which AI
changes how people work inside the firm, and a product channel, in which AI capabilities are
built into what the firm sells. Using text from product descriptions and job postings, we find
that embedding AI into the product, beyond layering on AI tools into existing workflows, is a
primary way startups are scaling “knowledge work” without large teams of knowledge workers.

∗
Kim: h@insead.edu. Koning: rem@hbs.edu. The authors gratefully acknowledge financial support from the AI
Institute at Harvard Business School and INSEAD. The authors thank Samay Gupta and Simonne Pinto for research
assistance. We thank Ajay Agrawal, Arnaldo Camuffo, Lauren Thomas, Todd Lensman, and Abhishek Nagaraj along
with seminar participants at SMS, the HBS EM working lunch, and the INSEAD brownbag for their especially helpful
feedback.

1

1

Introduction

Technology reshapes organizational form. Railroads and the telegraph allowed firms to serve national markets, increasing the returns to coordinating across distance and creating the conditions
for large managerial hierarchies (Chandler, 1993). Electricity altered how work could be organized
on the factory floor and spurred the rise of scientific management (David, 1990). The internet
lowered communication and coordination costs, supporting more platform-based, networked, and
distributed forms of organization (Shapiro and Varian, 1999). Across technologies—from lithography to Linux—product and organizational architectures often co-evolve (Henderson and Clark,
1990; Colfer and Baldwin, 2016).
How will AI reshape firms? Recent research has largely focused on a process channel, emphasizing how AI tools augment or automate work inside the firm through agentic coding, AI-assisted
customer service, automated sales operations, and many other internal tasks (Peng et al., 2023;
Brynjolfsson, Li, and Raymond, 2023; Dell’Acqua et al., 2023; Csaszar, Ketkar, and Kim, 2024;
Otis et al., 2025; Sarkar, 2026; Kim, Kim, and Koning, 2026). Yet AI can be used for more than
internal work. AI-native firms embed AI into their products, allowing customers to generate slide
decks with AI, build software, and even run their accounts receivable. This product channel shifts
productive capability from the internal organization into the product, moving knowledge work that
would have required large teams to be performed directly in the product itself through imported
model capabilities. Both channels suggest that “AI-native” firms—those that are built around AIenabled products and processes—may be organized differently than comparable non-AI firms.
In this paper, we highlight this product channel and explore whether AI-native firms are, in
fact, organized differently. We focus on startups because new firms are less constrained by legacy
structures, making them a likely setting in which new organizational forms first become visible.
Measuring whether a firm is AI-native presents a challenge because, as our two channels illustrate,
there are many ways companies can be built around AI. Here, we leverage the fact that Y Combinator (YC), a leading accelerator for early-stage companies, curates and encourages startups to tag
themselves with public labels to help investors, journalists, and customers discover new ventures
they might be interested in. We leverage the fact that startups can add an “AI” tag to their public
profile, enabling us to identify startups that claim that AI is central to their firm, whether that is
embedding it into their products, using it as a core part of their production process, or something
else entirely. To test if our YC results generalize beyond this extremely selected set of firms, we also

2