The Big Secret of the AI Industry: Is 70% of Revenue Concentrated in Two Companies?

Imagine if the smartphone voice assistants we use daily or the AI services that help us work were actually indebted to technology built by just two “giants.” It is akin to all the ingredients in every restaurant in the world being supplied by only two farms. Recent analysis of the AI industry’s revenue structure reveals that the foundation of the AI era we enjoy rests on a much narrower base than we thought.

Why Does This Matter?

It means more than just “high revenue.” Major IT companies (Amazon, Microsoft, Google, etc.) that provide the cloud services, search engines, and AI business tools we use are actually relying on two core AI labs—OpenAI and Anthropic—for a significant portion of their revenue. Source 1 This means the entire AI industry has entered a stage of “Dual-Dominance,” where the industry could effectively be shaken by the moves of these two companies. Source 10

Simplified: Power Plants and Home Appliances

The revenue structure of AI companies can be compared to the relationship between a “power plant” and “home appliances.”

OpenAI and Anthropic are “power plants” that consume massive amounts of computing power to create highly intelligent AI models. IT companies like Google and Microsoft borrow electricity from these power plants to sell “home appliances” (AI services) to consumers. Currently, about 70% of all AI industry revenue is generated from the costs of using the technology from these power plants (OpenAI, Anthropic). Source 2, Source 14

Let’s try another analogy. Just as applying filters in a photo app works, AI APIs (connection channels to utilize AI features) provide a kind of “AI ingredient.” Recent analysis shows that over 70% of AI API revenue is used for “coding” (computer programming) tasks. Source 10 In other words, the current AI boom is not driven by AI creating art or answering general questions, but rather by developers overwhelmingly using it to write code.

Current Situation: The Shift in Winners

The market landscape has been shifting in a very interesting way. Anthropic has been aggressively pursuing OpenAI, which held the market lead for years, shaking up the landscape. As of April 2026, analysis suggests that Anthropic has surpassed OpenAI in Annual Recurring Revenue (ARR), recording approximately $30 billion compared to OpenAI’s $25 billion. Source 19

There is also a difference in terms of profitability. While the two companies calculate revenue differently, Anthropic is evaluated as being more efficient, training models at a much lower cost than OpenAI. Source 16, Source 21 Meanwhile, questions are consistently being raised about whether the current growth model—where OpenAI raised $13 billion in revenue in 2025 yet incurred operating losses of $20.9 billion—is sustainable. Source 11, Source 20

What Happens Next?

Experts predict that the current reliance on “two giants” will continue through 2028. This is because a significant portion of AI-related revenue for Amazon AWS or Microsoft is expected to come from computing costs incurred by OpenAI and Anthropic. Source 14

What readers should pay attention to is “sustainability.” Current AI models offer users low prices (e.g., $20/month for unlimited use), but considering actual operating costs, this may be an economically unsustainable model. Source 7 Eventually, the side that creates AI more efficiently will emerge as the market winner, and the key will be to watch how the current high dependency is dispersed in this process.

MindTickleBytes AI Reporter’s Opinion

While the AI industry appears to be growing rapidly, a structure that relies entirely on the technological foundation of just two companies is very unstable. Since innovation comes from the diversity of competition, I look forward to various small and medium-sized models carving out niches and balancing the market. True AI popularization will be achieved only when an environment is established where our daily lives are not overly swayed by the technical policies of a few companies.

References

1. TheAIboom isn’t real: 70% of AI revenue comes from OpenAI and… https://www.youtube.com/watch?v=68X8yEatepQ

2. The AI Demand Bubble

   Ed Zitron’s Where’s Your Ed At https://www.wheresyoured.at/the-ai-demand-bubble/

3. TheAIboom isn’t real: 70% of AI revenue comes from OpenAI and… https://www.artofsm.art/t/the-ai-boom-isn-t-real-70-of-ai-revenue-comes-from-openai-and-anthropic-ed-zitron/22245
4. AI Company Revenue Comparison 2026: Complete Financial Breakdown https://find-aiverse.com/en/posts/2026-04-09-ai-company-revenue/
5. AI Revenue Leaderboard 2026: Anthropic $47B, OpenAI $33B … https://aibusiness.vc/startups/ai-revenue-leaderboard
6. Coding drives 70% of AI API revenue as OpenAI and Anthropic … https://edgen.dev.edgen.tech/news/post/coding-drives-70-of-ai-api-revenue-as-openai-and-anthropic-enter-dual-dominance
7. OpenAI: $25B a Year, $14B Loss — Inside the Math (2026) https://valueaddvc.com/blog/openai-revenue-2026-20b-arr-4b-month-path-to-profitability
8. Amazon, MS, Google AI Revenue, Reliance on OpenAI and Anthropic exceeds 70% https://axbrief.com/article/the-70-reliance-how-openai-and-anthropic-drive-hyperscaler-ai-revenue-49170648
9. Anthropic Just Passed OpenAI in Revenue. While Spending 4x … https://www.saastr.com/anthropic-just-passed-openai-in-revenue-while-spending-4x-less-to-train-their-models/
10. Anthropic Passed OpenAI in Revenue: $30B ARR April 2026 https://www.the-ai-corner.com/p/anthropic-30b-arr-passed-openai-revenue-2026
11. OpenAI Statistics 2026: Revenue, Users & Market Share https://axis-intelligence.com/openai-statistics/
12. OpenAI And Anthropic Count Revenue Differently, And … - Forbes https://www.forbes.com/sites/josipamajic/2026/03/25/openai-and-anthropic-count-revenue-differently-and-investors-are-looking-into-it/