On September 14, Nadella posted on X, explicitly supporting "consciously controlling the pace of R&D" to advance alignment research, and emphasizing that the pursuit of superintelligence must be centered on the core principle that "AI can help and be controlled by humans." He pointed out that the benefits of AI must be widely distributed and not controlled by a few entities. This move by Microsoft marks the first time it has publicly documented the behavioral standards for its first-party AI models. Previously, Anthropic, OpenAI, and Google had been discussing collaborating to form an AI industry standards body, a move supported by OpenAI CEO Altman.