What happened

AI made headlines this week for its name, its safety promises, and its business model. President Donald Trump signed an executive order telling US government departments to use “super intelligence,” or “SI,” instead of “artificial intelligence.” Around the same time, major technology leaders signed a voluntary AI safety pledge. It says advanced AI systems should behave as intended. It also calls for problems to be found and fixed quickly. Trump called the pledge “morally binding.” The document does not set clear legal penalties for breaking it. TechCrunch’s podcast summary describes these moves as part of the same week in AI.

The background

The policy news came as companies are giving AI friendlier faces. Meta, the company behind Facebook and Instagram, has Muse. OpenAI, the company behind ChatGPT, launched Dots. Both products were presented with friendlier personalities or cartoon-like identities. The goal is to make advanced software feel easier to approach. The larger commercial question is whether people will keep paying.

The same coverage points to an uncomfortable number. A market report cited in TechCrunch’s consumer AI analysis found that 2.2% of consumers paid for AI services in May. Their average spending was $31 per month. A Bank of America estimate put paid use among US consumers at roughly 3% in March. These are different surveys, so they are not one precise global rate. Still, they suggest that paying consumers remain a small share.

Why it matters

AI is unusually expensive to operate, the analysis argues. A large audience does not automatically create a profitable service. If the number of paying users stays small, revenue may not cover the cost of running advanced models. That helps explain the industry’s shift toward enterprise customers. Enterprise customers are businesses buying AI for work. OpenAI’s Dots materials also mention software developers and creative agencies. A personal assistant can therefore become a route into company sales. That is a strategy, not proof that consumer AI is already a healthy business.

What we can confirm

Three facts are clear from this reporting. The US government changed its official wording. AI leaders published a voluntary safety promise. Survey estimates put paying consumer use near 2% to 3%. Those facts describe adoption and policy language. They do not prove that the new label improves safety or that a product makes money.

What remains unknown

The 2.2% figure measures payment, not every kind of use. It does not explain why people do not pay. It does not show whether $31 covers a company’s operating costs. The safety pledge may create a shared expectation, but this material does not show how it will be enforced. We also do not know whether friendly products will produce lasting subscriptions.

What to watch next

Watch four numbers: paying users, average monthly spending, operating costs, and enterprise bookings. Watch whether Muse and Dots keep users after their novelty fades. On safety, look for published tests, incident reports, and independent oversight. The central question is simple. Will the bigger label change how AI is governed and paid for, or only how it is presented?