Microsoft proposes an AI “constitution”

Microsoft has unveiled a Code of Conduct for artificial intelligence that it treats as a kind of “constitution” for the development and use of future AI systems.

The document aims to set clear boundaries for artificial systems and guarantee that ultimate control remains in human hands.

According to Microsoft, AI models must be designed to accept human corrections and interventions and never resist being shut down. The company also requires systems to be understandable and predictable, so users can understand how and why they make decisions.

The code also sets limits on the use of artificial intelligence in dangerous activities. Microsoft aims to prevent models from being used to develop weapons, cyberattacks, or other activities that could cause harm.

One of the document’s key principles is that AI must not take control over people. According to the company, artificial systems can become increasingly powerful and autonomous, but they must always remain under human oversight and authority.

Microsoft also does not consider artificial intelligence a conscious being and does not support the idea that AI systems should have rights or legal status similar to humans.

The document was prepared over recent months, and Microsoft has opened a six-week public consultation period. The company has called on experts and the public to share their views and suggestions on the proposed rules.

The initiative comes at a time when the rapid development of artificial intelligence has raised concerns about safety, autonomy, and the possibility that advanced systems could act in unpredictable ways. Leading technology companies are facing growing pressure to put stronger control mechanisms in place.

For Microsoft, the core message is clear: the more powerful artificial intelligence becomes, the more important human-set boundaries are. The company aims for the Code of Conduct to serve as the foundation for how future AI systems are developed and controlled.