WorkSeptember 14, 2026via The Decoder
Microsoft's AI rulebook: readable thinking, no inner life, and definitely no rights
Why it matters
Microsoft is staking out a policy position on AI safety and autonomy that differs from competitors like Anthropic. For practitioners building enterprise AI systems, this signals how major vendors will frame responsibility and control — directly affecting procurement, compliance, and deployment strategy.
Key signals
- Microsoft publishes code of conduct for MAI models emphasizing human control over autonomy
- Mustafa Suleyman quote: 'If it isn't safe we shouldn't build it'
- Microsoft explicitly rejects artificial inner life or consciousness claims for models
- Positioned as contrast with Anthropic's approach to model agency/consciousness
- Microsoft published a code of conduct for MAI models
- Mustafa Suleyman (AI chief) quoted: 'If it isn't safe we shouldn't build it'
- Policy rejects artificial inner life or consciousness claims for models
- Explicit contrast with Anthropic's approach
- Human control positioned ahead of autonomy and performance
The hook
Microsoft's AI rulebook rejects consciousness claims and prioritizes human control — a philosophical stance with real implications for how enterprises think about agent deployment.
Microsoft AI has published a code of conduct for its MAI models that puts human control ahead of autonomy and performance. "If it isn’t safe we shouldn’t build it.," says AI chief Mustafa Suleyman. Unlike Anthropic, Microsoft rejects any form of artificial inner life or claims to consciousness for i…