2025年一季度AI&ML公共报表和估值指南2025

EMERGING TECH RESEARCH AI & ML PublicComp Sheet andValuation Guide Key takeaways PitchBook Data, Inc. Nizar TarhuniExecutive Vice President of Researchand Market Intelligence Paul CondraGlobal Head of Private Markets Research Hyperscalers will likely continue strong AI investment, despite any recession:As of April 4, the stocks of hyperscalers Amazon,Google, and Microsoft were down 21.3%, 18.6%, and 10%, respectively, from the recent stock market high of February 19. AI showsenough promise that the major hyperscalers will likely see any VC slowdown as an opportunity to maintain large investments in AI. James UlanDirector of EmergingTechnology Research Institutional Research Group Analysis Medium and large customers of hyperscaler AI infrastructure will likely modestly reduce spending on AI services, as most companyleaders instinctively cut spending in the face of a recession. Furthermore, AI has a cost. Building out AI-powered tools and workflowsrequires new software developers and sometimes new budget. Some leadership teams might be hesitant to make these investmentseven if they save money in the long term. However, in a softer environment, firms that maintain or accelerate their focus on AIshould do well. James UlanDirector of EmergingTechnology Researchjames.ulan@pitchbook.com pbinstitutionalresearch@pitchbook.comPublished on April 11, 2025 Key takeaways2Stock returns4Valuations5Revenue6EBITDA8 AI is likely to progress nicely:Despite an increasing likelihood of a recession, we expect capital to continue flowing into major VC-backed AI companies (such as OpenAI and Anthropic) given the power of the technology and their backing from capital-rich techcompanies, including Amazon, Microsoft, and Google. In a weaker economic environment, outside of the blue-chip AI companies such as OpenAI and Anthropic, VCs will likely favorAI companies with revenue, product leadership, and incredible founders. As always in venture, capital flows more easily to topfounders, such as Mira Murati, Elon Musk, and Ilya Skutskever, even if they have yet to generate revenue. In a recession, pre-seed and seed check sizes would decrease, making it harder for certain AI startups to raise a full pre-seed orseed round. This would apply to all startup categories. A severe recession would create a harsher fundraising environment. In suchscenarios, capital generally dries up, which would provide an advantage to Amazon, Microsoft, and Meta, as they can subsidizetheir AI services with cash flows from their core businesses. Finally, we continue to watch DeepSeek and the impact of other high-powered, low-cost models, as some enterprises have begun to use DeepSeek. PitchBook clients can accessthefull Excel data packfor thisreport via the Details tab in thedocument viewer. Components and parts sourcing for datacenters are a focus:One big question that remains is whether increased tariffs on Chinaand the East will slow the delivery of critical datacenter components to the US or decrease demand to build datacenters by drivingup cost. Morningstar analyst Eric Compton recently wrote about increased prices, saying, “Hardware supply chains are concentratedin Asia, with notable concentrations in China (34% additional tariff rate), Taiwan (32%), South Korea (25%), Japan (24%), Singapore(10%), Malaysia (24%), Thailand (36%), and the Philippines (17%). … Under the current executive order, semiconductors are exempt,and only hard goods are being tariffed.”1It is feasible that Eastern suppliers of datacenter components could experience financialstrain as tariffs make their components more expensive, thus reducing demand for datacenter components and eating into companyprofit margins. This comp sheet identifies public companies with significant revenue from AI-centric software and AI-focused products in both hardwareand software. Companies with diverse business models and AI leadership are referred to as “conglomerates,” while companies whoseproducts are centered on AI systems are labeled as “pure-play.” This qualitative categorization is not exhaustive yet can offer a directionalcomparison for private companies. The PitchBook AI & ML comp sheet was constructed with the PitchBook Excel plugin utilizing both PitchBook and Morningstar data. Thetool allows subscribers to pull financial data and company information into Excel for over 55,000 public companies across the US and theworld, as well as PitchBook’s proprietary data on over 4 million private companies. Stock returns Valuations Revenue Revenue EBITDA EBITDA ©2025 by PitchBook Data, Inc. All rights reserved. No part of this publication may be reproduced in any form or by any means—graphic, electronic, or mechanical,including photocopying, recording, taping, and information storage and retrieval systems—without the express written permission of PitchBook Data, Inc. Contents arebased on information from sources believed to be reliable, but accuracy and completeness cannot be guarant