U.S. AI Startups Face Scrutiny Over Inflated Revenue Figures

Firms Exploit Gaps in Mixed Subscription and Usage-Based Billing One-Month Revenue Spikes Projected Across a Full Year "Everyone Else Does It" Defense Draws Growing Criticism

International|
|
By Cho Yang-joonmryesandno@sedaily.com
||

The annualized method that artificial intelligence startups in Silicon Valley have adopted to measure revenue is drawing scrutiny as some of them prepare to go public. The startup industry argues the approach is unavoidable if companies want to show growth, but critics say it can overstate actual performance. Anthropic and OpenAI have also used this method to convince investors of their growth, and the practice is likely to be tested as they move toward listings.

Business Insider reported on the 8th that AI startups are increasingly interpreting or distorting revenue metrics at their own discretion, a phenomenon known as revenue inflation. The measure AI startups typically use is annual recurring revenue, or ARR, which estimates a full year of performance based on revenue expected to repeat under subscription contracts signed with customers. It is used mainly by software companies.

- - Seoul Economic Daily International News from South Korea
-

The problem is that AI startups charge not only monthly subscription fees but also usage-based fees tied to tokens or hours of use. In OpenAI's case, ChatGPT subscription fees, which are set monthly for individual customers, can be counted as recurring revenue. But revenue from usage-based application programming interface, or API, fees, which are mainly used by corporate and developer customers, and revenue from its advertising business are difficult to calculate as recurring.

As a result, a growing number of AI startups are applying the run rate method, which projects a full year of performance based on results from a specific period. The approach varies by company. OpenAI calculates ARR based on subscription revenue while measuring its newly launched advertising business on a run rate basis. Anthropic, by contrast, applies run rate to its entire revenue.

Some startups, however, are abusing the method by announcing run rate figures based on a month in which they posted what amounts to a one-off revenue spike, distorting their performance. Akshay Nariseti, chief executive of Pocket, a maker of AI recording devices, said there is no guarantee that a sudden jump in revenue will continue for a full year.

OpenAI and Anthropic, the most prominent AI startups, are no exception. OpenAI calculates ARR based on subscription revenue while measuring its newly launched advertising business on a run rate basis. Anthropic applies run rate to its entire revenue.

Some companies have gone further and simply padded their revenue figures. Roy Lee, chief executive of AI startup Cluely, told media last year that the company's ARR was $7 million (about 9.4 billion won). After the claim drew controversy, Lee acknowledged eight months later that the statement had been blatantly dishonest and that the actual figure was $5.2 million. But when startup executives are pressed about figures that appear inflated, they often respond with the excuse that everyone else does the same thing, Business Insider said.

In the venture capital industry, some are warning, with a note of self-deprecation, that such unethical behavior by startups mirrors market conditions in late 2021, when the venture bubble was at its peak.

Original reporting by Cho Yang-joon for Seoul Economic Daily.

AI-translated from Korean. Quotes from foreign sources are based on Korean-language reports and may not reflect exact original wording.

Watch · Seoul Economic Daily

More →
3:12

AI KEY

Preview
Korean Corporate Intelligence HubKOSPI · KOSDAQ · 12 sectors

A live, cap-weighted view of every KOSPI and KOSDAQ sector, with same-day Korean reporting distilled by company — built for foreign investors, correspondents and analysts who need to scan Korea before the next session.

Korea Company Atlas

Preview
Market Ontology · The Feedback LoopKFTC 2025 · 92 groups · 121,954 articles

An English ontology of the Korean market — how companies, the media, the government and the National Assembly move each other in a loop. Korea's named controlling persons and designated business groups are a mechanism, not a risk to be priced blind.

SIGNAL

Now live
English Edition · Capital MarketsM&A · IPO · PE · Fund Flows

SIGNAL English Edition is live — Korea's deal desk reporting in English. M&A, IPOs, private equity and fund flows, covered daily for global institutional investors. Browse free; subscriber-only scoops at the 50% intro rate.