Article from TechCrunch investigates concerns among founders, investors, and finance professionals about the misuse and inflation of annual recurring revenue (ARR) metrics in the AI startup ecosystem.

It highlights allegations that some AI companies are overstating revenue by substituting “contracted ARR” or “committed ARR” (CARR) for true ARR, or by using annualized run-rate revenue based on short-term performance.

These practices can significantly exaggerate a company’s financial health, especially when long-term contracts, pilot programs, or usage-based billing are involved.

Several sources cited in the report claim that some startups count signed contracts or even unpaid pilot programs as ARR, despite uncertainty about whether the revenue will ever be realized.

In some cases, investors are aware of these adjustments but tolerate them due to competitive pressures and the rapid growth narrative surrounding AI companies.One VC noted that companies may report CARR significantly higher than actual ARR, inflating perceived growth.There are also examples where startups publicly claim tens or even hundreds of millions in ARR while only a portion of that is from paying customers.

The article explains that ARR was originally intended to reflect stable, recurring revenue from contracted customers, but it is not governed by strict accounting standards like GAAP.This leaves room for interpretation and manipulation.

The rise of AI startups and heightened investor expectations for exponential growth have intensified pressure to present strong metrics, sometimes at the expense of accuracy.Ultimately, the report suggests a tension between storytelling and financial transparency in venture-backed AI startups.

While some founders defend strict, honest reporting practices, others argue that the competitive environment incentivizes aggressive metric interpretation.Critics warn that such practices could distort valuations and create misleading signals for talent, investors, and the broader market.