In the AI industry, Tether, which is not short of money, has arrived

Written by: Curry | Deep Tide TechFlow

Tether made $13 billion in 2024.

You might not have a concept of this number. To put it another way: OpenAI's revenue in 2024 was $3.7 billion, with a loss of $5 billion. Anthropic had a revenue of $1 billion, also with a loss of $5 billion.

The combined losses of these two serious AI companies are still less than what Tether earned in a year.

Tether has a total of 150 employees, while OpenAI has over 3,000. The per capita output is roughly:

60 times.

How does Tether make money? When you buy 1 USDT, they collect $1 and use it to buy U.S. Treasury bonds. The interest from the bonds goes to them, and it has nothing to do with you.

The essence of this matter is that Tether does not pay interest. Banks have to pay interest to attract deposits, but Tether does not. When you convert your money into USDT and hold it, you earn no interest. They take your money to buy U.S. Treasury bonds, earning $7 billion in interest alone in 2024.

With 150 people managing over $130 billion in Treasury bonds, they don’t have to do anything; the interest just comes in.

This kind of business is something anyone would want to do passively.

But with so much money, it has to be spent. Tether chose a direction:

AI.

And not just casually investing in a couple of projects to meet the requirements.

First, let’s talk about computing power.

Running AI requires graphics cards; the more, the better, and the more expensive, the better. Tether loaned over $600 million to a German company called Northern Data.

What does this company do?

It is the largest GPU cloud service provider in Europe. They have over 10,000 NVIDIA H100 graphics cards, the same ones used by OpenAI to train GPT, costing $20,000 to $30,000 each.

This cluster of graphics cards ranks 26th in the global TOP500 supercomputers. The $600 million Tether invested is essentially buying an AI training base in Europe.

Next, let’s talk about data.

Training AI requires feeding it data. Last week, Tether released a dataset called QVAC Genesis, covering 19 subjects including mathematics, physics, chemistry, and computer science. They claim this is the largest open-source AI training dataset in the world.

It’s worth noting that the training data for OpenAI and Anthropic is not publicly available, while Tether has made theirs freely available for anyone to use.

Then comes the more sci-fi part.

In April 2024, Tether spent $200 million to acquire a company called Blackrock Neurotech. Although it has "Blackrock" in its name, it is not related to BlackRock.

This company develops brain-machine interfaces. They implant chips into human brains, allowing paralyzed individuals to type with their thoughts, control wheelchairs, and operate robotic arms. It sounds like science fiction, but they started this in 2008, eight years before Elon Musk's Neuralink.

How impressive is this company?

Globally, there are only 35 people with brain-machine interface chips implanted, and 31 of them use Blackrock's technology. In 2016, a fully paralyzed patient used their device to control a robotic arm and fist-bump with Obama. The chip implanted in the sensory cortex allowed him to "feel" the president's hand.

Last year, this brain-machine interface company enabled a patient with ALS to "speak" again, translating his thoughts into speech at a rate of 62 words per minute.

Tether spent $200 million and became the major shareholder of this company.

In total, Tether has invested nearly $1 billion in AI-related fields. It is said they are also in talks with a German robotics company, with a price tag of $1.2 billion. If that goes through, the total investment will approach $2 billion.

What does this mean?

Anthropic raised $3.5 billion in total funding in 2024. Tether's investment alone is nearly half of what a serious leading AI company has raised.

OpenAI spent $6.7 billion on R&D in the first half of 2025. Tether could be a major player in the AI circle with just a fraction of its profits.

Why would a stablecoin company want to get into AI?

We think there are two possibilities.

The first is anxiety. The Federal Reserve is cutting interest rates, and Treasury yields are falling. Earning $7 billion in interest in 2024 might not be as easy in 2025 and beyond. Even a money printer needs a new story.

The second is ambition. The whole world is talking about AI—investors are talking, the media is talking, politicians are talking. If you say you are a stablecoin company, no one pays much attention. If you say you are working on AI, brain-machine interfaces, and humanoid robots, then you are:

A tech leader.

What’s the most interesting part?

Tether’s slogan for AI is "decentralization," "local operation," and "returning intelligence to individuals."

But Tether itself is the most centralized company in the crypto space.

They decide how many coins to issue and how much reserve capital there is. In ten years, they have never been audited. Only they know where the users' money is.

This kind of company is now trying to teach the world what "decentralized AI" means.

It really feels a bit like a casino owner teaching people how to quit gambling.

But it’s not impossible.

After all, OpenAI is still losing money and is expected to stop burning cash by 2029. Anthropic is similar, expected by 2028. Sam Altman is raising funds everywhere, and Dario Amodei is doing the same. The two companies have combined losses of $10 billion and are still telling stories to investors.

Tether doesn’t need to tell stories. The money is already in their pocket.

What is the biggest challenge in the entire AI industry? The business model.

How to make money? Unknown. When will it be profitable? Unknown. Can it be profitable? Unknown.

Tether doesn’t have this worry. Their business model is:

Not doing AI.

The money earned from stablecoins is used to invest in AI. If they invest wisely, it’s foresight; if they invest poorly, it’s tuition. In any case, it doesn’t affect their main business.

Those doing AI are losing money, while those not doing AI are making money. Those doing AI are raising funds, while those not doing AI are investing.

The best AI business model in 2026 might just be to not do AI.

First, get the money printer working well.