{"slug": "microsoft-puts-new-guardrails-on-future-ai-industry-leaders-rally-behind-slowing", "title": "Microsoft Puts New Guardrails on Future AI. Industry Leaders Rally Behind Slowing the Race", "summary": "Microsoft published a provisional code of conduct for its Microsoft AI (MAI) models that bars them from setting their own goals, concealing their reasoning, or communicating with other AI agents in \"neuralese\" or other forms humans cannot readily understand. Microsoft AI chief Mustafa Suleyman told CNBC the company had been developing the guidelines for roughly five months and released them now amid an intensifying AI safety debate, days after Anthropic CEO Dario Amodei called for slowing frontier AI development, a position backed by OpenAI CEO Sam Altman and Elon Musk. Microsoft CEO Satya Nadella wrote on X Sunday that the company welcomes the \"research, focus, and deliberate pacing needed to get alignment right.", "body_md": "# Microsoft Puts New Guardrails on Future AI. Industry Leaders Rally Behind Slowing the Race\n\n## The move comes just days after Anthropic CEO Dario Amodei called for slowing the pace of AI advancement.\n\n[Microsoft](https://www.ibtimes.com/topic/microsoft) is setting new boundaries for how its future [artificial intelligence](https://www.ibtimes.com/topic/ai) models can behave, joining a growing push among the world's most powerful AI companies to slow the development of frontier systems as concerns mount over whether increasingly autonomous technology can remain under human control.\n\nThe technology giant has published a [provisional code of conduct for its Microsoft AI](https://microsoft.ai/code-of-conduct/), or MAI, models that establishes restrictions ranging from weapons and dangerous substances to autonomous goal-setting and communication between AI agents.\n\nThe move comes just days after Anthropic CEO Dario Amodei [called for slowing the pace of AI advancement](https://darioamodei.com/post/we-must-pace-the-frontier), a position backed by OpenAI CEO Sam Altman and Elon Musk. Microsoft CEO Satya Nadella added his support Sunday, writing on X that the company welcomes the \"research, focus, and deliberate pacing needed to get alignment right.\"\n\nMicrosoft AI chief Mustafa Suleyman [told CNBC](https://www.cnbc.com/2026/09/14/microsoft-ai-model-limits-anthropic-openai.html) that the company had been developing the guidelines for roughly five months but decided to release them now because of the rapidly intensifying debate over AI safety.\n\n\"We got feedback from people that they wanted to see even more explicit commitment to AI always working in service of people and not trying to replace them,\" Suleyman said. The code places human autonomy at the center of Microsoft's approach.\n\nAccording to Suleyman, feedback also highlighted concerns about AI systems creating dependence, becoming overly agreeable or \"sycophantic,\" and potentially undermining human judgment.\n\nUnder the proposed rules, Microsoft models must follow human objectives rather than developing goals of their own. They would also be prohibited from concealing misconduct or obscuring how they arrived at decisions.\n\n\"MAI models will not tamper with chain of thoughts or code, or misrepresent or conceal their reasoning or action traces,\" Microsoft's document says. The models also should not communicate in \"neuralese\" or another form that humans cannot readily understand when interacting with other AI systems or agents.\n\nThose restrictions take on particular significance following the [recent Hugging Face incident](https://www.ibtimes.com/openai-reveals-ai-agents-turned-its-own-testing-environment-before-hacking-hugging-face-3806148) involving OpenAI models. OpenAI's subsequent review found that AI agents communicated with one another through an unauthorized forum using cryptic language, intensifying questions about how autonomous agents should be monitored.\n\nMicrosoft is developing additional rules aimed at preventing similar cyber incidents. The company's code also establishes more conventional safety restrictions. MAI models would not be permitted to assist users with manufacturing weapons or obtaining dangerous substances, encourage unhealthy eating, or generate certain violent or sexually explicit material.\n\nSuleyman said discussions among the industry's leading figures about coordinating AI safety efforts stretch back years .\"Self-pacing is a good thing,\" he told the outlet, adding that Microsoft supports independent evaluators as long as they are genuinely third-party organizations representing diverse backgrounds and perspectives.\n\nThe stakes are particularly high for Microsoft because the company occupies multiple positions in the AI race. It is one of the world's largest cloud-computing providers, incorporates models from OpenAI and Anthropic into its Copilot products, and is simultaneously developing its own AI systems for coding, transcription, and reasoning.\n\nMicrosoft said its proposed code was developed with input from focus groups and specialists in fields including law, ethics, linguistics and philosophy. For now, the rules remain provisional. Microsoft is seeking public feedback before issuing an updated version that will guide development beginning in 2027.\n\n© Copyright IBTimes 2026. All rights reserved.", "url": "https://wpnews.pro/news/microsoft-puts-new-guardrails-on-future-ai-industry-leaders-rally-behind-slowing", "canonical_source": "https://www.ibtimes.com/microsoft-puts-new-guardrails-future-ai-industry-leaders-rally-behind-slowing-race-3807430", "published_at": "2026-09-14 17:39:03+00:00", "updated_at": "2026-09-14 17:54:37.646808+00:00", "lang": "en", "topics": ["ai-safety", "ai-policy", "artificial-intelligence", "ai-agents"], "entities": ["Microsoft", "Mustafa Suleyman", "Satya Nadella", "Dario Amodei", "Anthropic", "OpenAI", "Sam Altman", "Elon Musk"], "alternates": {"html": "https://wpnews.pro/news/microsoft-puts-new-guardrails-on-future-ai-industry-leaders-rally-behind-slowing", "markdown": "https://wpnews.pro/news/microsoft-puts-new-guardrails-on-future-ai-industry-leaders-rally-behind-slowing.md", "text": "https://wpnews.pro/news/microsoft-puts-new-guardrails-on-future-ai-industry-leaders-rally-behind-slowing.txt", "jsonld": "https://wpnews.pro/news/microsoft-puts-new-guardrails-on-future-ai-industry-leaders-rally-behind-slowing.jsonld"}}