Editorial illustration for Microsoft AI CEO calls Anthropic speculation 'really, really dangerous
Microsoft AI Chief Warns Against Anthropic Speculation
Microsoft AI CEO calls Anthropic speculation 'really, really dangerous
Microsoft published a 37-page document on Monday laying out what it calls a "humanist AI code of conduct," a direct response to the safety debate that's been building across the industry for weeks. The timing isn't accidental. Anthropic CEO Dario Amodei spent the weekend calling for a coordinated slowdown in AI development, echoing researcher warnings that model capabilities are advancing faster than anyone's ability to verify what these systems are actually doing once deployed as autonomous agents.
Microsoft's document stakes out a very different position than its rival's. It states flatly that "people matter more than AI," that models are not conscious, and that they "should not be designed to imitate consciousness." The company goes further, rejecting any push toward legal personhood for AI systems or the notion that models might deserve welfare protections or rights. That's a pointed rebuke of the AI welfare research Anthropic has been promoting, including Amodei's own comments earlier this year that he's "open to the idea" that models could be conscious.
Mustafa Suleyman, Microsoft's AI CEO, has already made his views on this clear in blunt terms.
Microsoft’s AI code of conduct makes it clear that “people matter more than AI,” and that AI models are not conscious and “should not be designed to imitate consciousness.”
Why this matters
Suleyman's "really, really dangerous" line isn't really about Anthropic's science. It's about who gets to set the terms of the safety conversation. Microsoft is trying to position itself as the reasonable adult in the room, rolling out a 37-page code of conduct while a rival CEO warns the industry might be moving too fast to control.
For developers and founders building on top of these platforms, that's worth watching closely: the vendor you pick is also picking your regulatory exposure and your PR posture when things go wrong. Suleyman himself admits Microsoft isn't yet a top-tier model provider, which makes the timing of this pushback notable. A company still chasing frontier status has real incentive to frame slowdown talk as fearmongering rather than engage with the substance.
None of this settles whether Amodei's concerns about verifying agent behavior are overblown or prescient. It does tell us that "safety" language is becoming a competitive weapon as much as a technical concern, and researchers should treat corporate safety branding with the same scrutiny they'd apply to a benchmark claim.
Related Reading
Browse all Anthropic news →
Research & Benchmarks Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks Databricks finds multi-step agents beat single-turn RAG by 21% to 38% on STaRK
Research & Benchmarks Nvidia's DLSS 4.5 beta adds 6x Multi Frame Generation for RTX 50 GPUs
Related Topic Claude gains shared context in Excel, PowerPoint; Microsoft adds Copilot Cowork
AI Tools & Apps
Related Topic Anthropic: Hackers Stole Claude Tokens Via Compromised Session Key
LLMs & Generative AI
From Archives Iris-mini and Iris-pro Lead Open-Weight Search Agent Class
From Archives UC Berkeley Law Bans AI From Graded Work in Skills Debate
From Archives Anthropic CEO says two factors convinced him to slow AI development
From Archives Anthropic CEO calls for industry-wide AI safety standards
Business & Startups
Common Questions Answered
What is Microsoft's 'humanist AI code of conduct' and why did the company release it?
Microsoft published a 37-page document outlining a humanist AI code of conduct as a direct response to the safety debate building across the AI industry. The code of conduct emphasizes that people matter more than AI and that AI models should not be designed to imitate consciousness, positioning Microsoft as a responsible player in the ongoing AI safety conversation.
Why did Microsoft's AI CEO call Anthropic's speculation 'really, really dangerous'?
Microsoft's AI CEO criticized Anthropic's calls for a coordinated slowdown in AI development, viewing it as dangerous speculation rather than legitimate science. The disagreement reflects a broader debate about who gets to set the terms of the AI safety conversation and establish industry standards for responsible development.
What safety concerns did Anthropic CEO Dario Amodei raise about current AI development?
Amodei warned that model capabilities are advancing faster than anyone's ability to verify what these systems are actually doing once deployed as autonomous agents. He called for a coordinated slowdown in AI development to address researcher concerns about the pace of advancement outpacing safety verification capabilities.
How might a developer's choice of AI vendor affect their regulatory position?
The vendor a developer chooses to build on top of is also picking their regulatory framework and positioning in the safety debate. As Microsoft and Anthropic take different stances on AI development speed and safety protocols, developers need to carefully consider which platform aligns with their values and regulatory expectations.
Further Reading
- Mustafa Suleyman, CEO of Microsoft AI, said it's 'really, really dangerous' to speculate about AI consciousness - The Verge
- Microsoft AI head calls out Anthropic for acting like Claude is conscious - The Verge
- Microsoft, Google, Amazon say Anthropic Claude remains available to customers except the Defense Department - TechCrunch
- Anthropic ditches its core safety promise in the middle of an AI safety debate - CNN
- Satya Nadella Backs Human Control Amid AI Safety Fears - Mathrubhumi
