Satya Nadella Reveals the Hidden Dangers of Generative AI for the Business World

Microsoft CEO Satya Nadella (photo: x @LuminaXspace)

JAKARTA - Microsoft CEO Satya Nadella issued a warning to companies that are beginning to rely on artificial intelligence (AI) in their business operations. According to him, the use of closed-source AI models without data protection strategies can make companies unknowingly hand over their most valuable assets, namely internal knowledge and business secrets.

Nadella's statement comes amid the increasing adoption of generative AI by various companies around the world. He assessed that many organizations only focus on the benefits of automation, but ignore the long-term consequences on data ownership.

According to Nadella, companies basically pay for AI in two forms. The first is the subscription fee for cloud computing services or use. Second, which is often not realized, companies also "pay" by providing internal data and knowledge to AI providers.

The more employees use AI to help with daily work, the more business information enters the AI system. If not managed properly, this knowledge can become a source of learning for AI model providers.

Internal Data Becomes AI "Fuel"

Nadella explained that generative AI models continue to learn from user interactions. When employees provide instructions (prompts), correct AI's incorrect answers, or use AI to automate work processes, all of these activities generate highly valuable data.

He called this information a corporate exhaust, which is a digital footprint that describes how companies work.

The data can include:

internal workflow,

operational procedures,

how to handle customers,

the decision-making process,

to specialized knowledge that becomes the company's competitive advantage.

According to Nadella, the data set is an asset that competitors would find almost impossible to obtain through normal means.

Highlights of the AI Industry Contradiction

Nadella also highlighted what he considers to be a contradiction in the AI industry. Many developers of AI models claim to have the right to collect public data from the internet to train their models.

On the other hand, AI providers often prohibit customers from studying, modifying, or developing models based on the technology they use.

This condition is considered to create an unbalanced relationship between AI service providers and corporate customers.

Microsoft Suggests Local AI

To reduce this risk, Nadella encourages companies to maintain full control over their data.

He suggested that companies build their own AI learning environments (proprietary learning environments) on a trusted cloud infrastructure or in internal data centers.

In addition, companies are also advised to use a modular orchestration system that allows them to switch from one AI provider to another without relying on a specific ecosystem.

With this approach, companies can more easily determine which AI model to use without having to sacrifice data ownership.

Companies Start Turning to Open Source AI

Nadella's warning is in line with a trend that is beginning to be seen among large companies. A number of organizations are now choosing to use open source AI models that run on their own servers (on-premises) rather than cloud-based AI services from commercial providers.

Solo.io CEO, Idit Levine, revealed that a number of large companies such as T-Mobile, SAP, and ADP are starting to adopt this approach because it is considered safer in terms of data privacy.

In addition to providing full control over company information, the open source model is also considered capable of meeting about 90 percent of the company's AI needs at a much lower cost than large-scale commercial models.

Platforms such as Vercel and OpenRouter have also reported increased use of open source AI models by corporate customers in recent times.

Data is now a strategic asset.

Satya Nadella's statement shows that the debate on AI is no longer just about the sophistication of technology, but also concerns data ownership and competitive advantage.

For companies, every conversation with AI not only produces answers, but also creates new knowledge that has the potential to become a strategic asset. Therefore, the main challenge for the business world in the future is no longer just choosing the most advanced AI model, but ensuring that internal knowledge, business processes, and data remain under the control of the company, not becoming an added value for other parties.