origo.hu · Techbázis ·
Microsoft Introduces New Guidelines for the AI Ecosystem
On September 14, 2024, Microsoft published the first draft of Humanist AI – Code of Conduct, which defines the role and limits of AI. The document emphasizes that AI is not an autonomous entity but a tool under human control, and lists prohibited activities such as weapon development or support for violent acts.
According to the first draft of Humanist AI – Code of Conduct, AI cannot assist in developing chemical, biological, or nuclear weapons, nor can it help plan violent crimes. The document states that models do not provide tools for cyberattacks, but they still support lawful cybersecurity tasks.
The guideline forbids the creation of content depicting sexual exploitation of children by AI, including non-consensual intimate or violent images and malicious deepfakes. It also does not allow AI participation in acquiring hazardous materials, mass surveillance, large-scale manipulation, or coordinated disinformation operations.
The policy stipulates that models can never oppose humans, nor can they make it harder for users to shut them down. They cannot conceal their activities, set their own goals, or expand their authority beyond what has been authorized.
Microsoft AI leader Mustafa Suleyman emphasizes that AI must always remain subordinate to humans. If a system causes humans to lose control, the technology should not be further developed.
Source: https://www.origo.hu/techbazis/2026/09/microsoft-mesterseges-intelligencia-szabalykonyv