One-Third of Major AI Companies Hide Their Numbers. That's a Problem for Buyers.
A new study of the 15 leading AI companies finds that only five regularly disclose concrete usage figures or revenue data - leaving enterprise buyers, regulators, and boards making vendor decisions on incomplete information.
The AI Company Comms Study 2026, released by 5WPR in partnership with Everything-PR, benchmarked how OpenAI, Anthropic, Google DeepMind, Microsoft, Meta, and 10 others communicate adoption and financial performance to the market. The findings expose a structural problem: disclosure practices vary wildly, and the companies generating the most revenue don't necessarily generate the most transparency.
Revenue Leaders Don't Always Lead in Disclosure
Anthropic's annualized revenue reached approximately $30 billion in April 2026, exceeding OpenAI's $25 billion as of February 2026. Yet OpenAI maintains a consumer user base roughly 20 times larger and receives several multiples more press coverage. Communications volume, the study concludes, no longer tracks with revenue volume.
OpenAI, Anthropic, Microsoft AI/Copilot, Google DeepMind/Gemini, and Perplexity rank as the most transparent on usage and revenue. The remainder publish aggregated numbers tied to funding rounds or disclose nothing systematically at all.
Midjourney, despite global brand recognition, publishes effectively no public revenue, user, or enterprise customer data. Character AI, Stability AI, Inflection AI, and Mistral operate with similarly sparse disclosure despite their scale.
Enterprise Penetration Is Now Measurable
The forecast era has ended. Anthropic discloses 8 of the Fortune 10 as customers. OpenAI is deployed at 92 percent of the Fortune 500. Microsoft Copilot has crossed 15 million paid seats inside its 400-million-plus Microsoft 365 installed base.
These numbers matter because they're concrete. A CIO evaluating vendors can point to them. Without them, procurement teams operate blind.
The Founder Voice Problem
Elon Musk, the most-covered AI founder by volume, registered 37 percent favorable sentiment against 55 percent unfavorable in Morning Consult tracking as of June 2025 - the lowest figures of his tracked series. In every other U.S. business sector, communications volume correlates with favorable sentiment. In AI, that correlation has inverted.
Sam Altman, Mark Zuckerberg, Dario Amodei, and Perplexity's Aravind Srinivas round out the loudest founder voices. Midjourney, Character AI, Stability AI, Inflection AI, and Mistral operate with minimal founder visibility relative to their scale.
What This Means for Communications Teams
The study identifies a sector-wide problem: asymmetric information. Every board evaluating AI vendors operates on different data sets. Some companies disclose monthly active users. Others don't. Some publish customer rosters. Others stay silent.
For PR and communications professionals, the implication is direct. Disclosure strategy now affects how enterprise buyers perceive vendor credibility. Transparency on usage, revenue, and customer adoption has become a competitive signal - not a compliance burden.
The study will be updated quarterly and will eventually incorporate primary polling of enterprise AI buyers, investor sentiment data, and coverage of open-source AI and international markets.
AI Learning Path for Public Relations Specialists covers how communications professionals can apply AI to brand strategy, media outreach, and reputation management - skills increasingly relevant as AI companies navigate disclosure decisions and founder positioning.
Your membership also unlocks:
