Хочете будувати кар’єру в Data Science та AI? Vodafone Big Data Lab відкриває новий набір —>
Наталя ХандусенкоAI Eng
14 July 2026, 09:34
2026-07-14
"You're essentially paying for intelligence twice," Microsoft CEO issues shocking warning to companies using AI, offers solution
There is growing concern in Silicon Valley that large AI labs that sell closed commercial models are acting as Trojan horses. Companies like OpenAI and Anthropic are increasingly gaining access to their customers’ most sensitive business information. They could then use that knowledge to their own advantage and potentially become competitors to those they serve.
Among those sounding the alarm are venture capitalists like Jason Calacanis and Palantir CEO Alex Karp, and they have recently been joined by Microsoft CEO Satya Nadella.
There is growing concern in Silicon Valley that large AI labs that sell closed commercial models are acting as Trojan horses. Companies like OpenAI and Anthropic are increasingly gaining access to their customers’ most sensitive business information. They could then use that knowledge to their own advantage and potentially become competitors to those they serve.
Among those sounding the alarm are venture capitalists like Jason Calacanis and Palantir CEO Alex Karp, and they have recently been joined by Microsoft CEO Satya Nadella.
What Microsoft CEO warns about
Satya Nadella warns that AI users (or “buyers,” as he calls them) are paying twice. They are knowingly spending money to use AI tokens, but at the same time, without realizing it, they are giving away their valuable data, writes TechCrunch.
“You essentially pay twice for intelligence: once in money, and once in something even more valuable: your own knowledge, which you must uncover to make that intelligence useful. And the better results you expect from a model, the more such knowledge you have to put into it!” he writes.
The most dangerous thing, in his opinion, is that companies literally teach models the intricacies of their business.
“Models learn from ‘emissions’—prompts that people write, tools that AI agents use, and especially corrections that users make when the model gets it wrong. Each such correction becomes institutional know-how,” he writes.
This is precisely "the knowledge that a competitor could never buy," and yet companies give it away themselves.
Nadella argues that if AI companies are free to scrape the internet to train their models, it would be only fair if businesses were given the opportunity to explore—or “distill”—these models in return.
“Distillation” is the practice of using the results of a model to understand how it works and train a new, usually cheaper model based on the data.
According to Nadella, modelers cannot sit on two chairs. It is hypocritical for them to freely train their systems on the world's data while prohibiting others from doing the same with their own models.
“While we need the great innovations that come from developers’ right to use public data in good faith to train models, I find it ironic that the current practice is to then turn around and impose strict restrictions on distillation,” Nadella writes.
Nadella is particularly concerned when modelers “reserve the right to learn from usage and customer interaction data.”
What is a possible solution?
Nadella’s solution is exactly what the CEO of a giant cloud provider would recommend. He wants companies to “retain ownership” of their data, including prompts, feedback, and so on. So he’s urging them to create their own “learning environments” in the cloud (where their data is likely already stored and, very conveniently, could very well be Microsoft’s cloud, Azure).
He also wants companies to implement what he calls “orchestration layers”—essentially a way to easily switch between AI models from different vendors, rather than being locked into one. Tools like AI gateways that allow companies to do just that are becoming increasingly popular.
While Nadella never uses the phrase “open source” as a way to retain ownership of data, that’s the obvious implication. But there’s another implication here.
Large companies, many of which still have their own data centers in addition to using the cloud, are already moving to open source models, installing them on their own local facilities.
Given that even the CEO of Microsoft—a company that invested in both OpenAI and Anthropic—is now openly urging businesses to be cautious about using closed models, it’s safe to predict that this trend will only grow. “When you consume intelligence, you create intelligence. And what you create should be yours,” Nadella notes.
“The point of potential collapse may be just around the corner.” Chinese hedge funds warn of AI “super bubble”: the market is overheated and at risk of collapse
Як нейромережі бачать вільну та незалежну Україну? Тест dev.ua
Нейронні мережі для генерації зображень бачать світ по-своєму, їхню логіку зрозуміти часом зовсім неможливо. Але таки хочеться. На честь Дня Незалежності України редакція dev.ua вирішила провести невеликий експеримент.
Ми задали чотирьом різним нейронним мережам п’ять однакових запитів: «прапор України», «День Незалежності України», «український Крим», «перемога України» та «українці». Отриманими результатами ми ділимося з вами нижче.
У TikTok тепер можна генерувати фон за допомогою нейромережі. Ми протестували її та ділимося результатами
У TikTok з’явилася нова функція «Розумний фон». З її допомогою як фон для тіктоків можна підставляти згенеровані нейромережею зображення. Редакція dev.ua протестувала цю технологію і ділиться своїми враженнями.
