Xiaohongshu Draws Red Lines for AI Content, Microsoft Ends Revenue Sharing with OpenAI

Industry News Roundup

The AI industry saw several major developments this week: Xiaohongshu systematically unveiled its AI content governance framework for the first time, drawing clear boundaries for creators; Microsoft and OpenAI's commercial partnership underwent a significant restructuring, with the revenue-sharing model officially coming to an end. These developments reflect the accelerating maturation of the AI industry in terms of content governance and business models.

Xiaohongshu Launches AI Governance Framework: Encouraging Creativity, Opposing Fabrication

On April 27, Xiaohongshu hosted its inaugural AI Governance Open Day in Beijing, publicly releasing the platform's systematic governance stance on AI content for the first time. This makes it one of the earliest mainstream content communities in China to articulate its AI governance position in such a comprehensive and public manner.

The core philosophy behind Xiaohongshu's governance framework can be summed up in one sentence: Encourage AI as a creativity amplifier, and firmly oppose AI being reduced to a tool for fabrication and low-quality content production.

Specifically, the platform divides its governance framework into two major categories: "Encouraged" and "Opposed." On the encouragement side, Xiaohongshu supports creators in leveraging AI technology to enhance creative quality and efficiency, unlocking greater creative possibilities. On the opposition side, the platform has explicitly outlined four categories of red-line behaviors:

• Using AI for unauthorized operations: Including mass-generating account content, manipulating traffic, and similar activities
• Using AI for fabrication: Creating deceptive content such as forged real-world scenarios and fabricated facts
• Using AI for infringement: Using others' likenesses, works, or other materials for AI generation without authorization
• Using AI for low-quality creation: Mass-producing homogeneous content with no real value

Particularly noteworthy is Xiaohongshu's "dual safeguard" approach to content labeling: All content created with AI involvement must be proactively labeled by creators, and for AI content that is not voluntarily labeled, the platform will use technical detection methods to identify and uniformly apply AI content labels. This mechanism balances creator self-regulation with platform oversight and is expected to create a reasonably effective closed-loop system in practice.

Against the backdrop of increasingly pervasive AI-generated content, Xiaohongshu's move carries significant industry-leading implications. As image and video generation technologies rapidly evolve, content platforms face an ever more daunting challenge of distinguishing authentic from fabricated content, making the establishment of clear rule systems an urgent industry necessity.

Microsoft Ends Revenue Sharing with OpenAI, Partnership Enters New Phase

Meanwhile, across the Pacific, the AI business landscape is also being quietly reshaped. Microsoft and OpenAI recently announced the next phase of their partnership, with the most notable change being: Microsoft will no longer pay revenue share to OpenAI.

Previously, Microsoft and OpenAI maintained a complex revenue-sharing mechanism under which Microsoft distributed OpenAI's model capabilities through its Azure cloud services and paid OpenAI a share of the revenue at agreed-upon rates. This adjustment signifies that the commercial relationship between the two parties is gradually transitioning from an early model of deep interdependence to a more independent and flexible partnership.

This shift comes against the backdrop of OpenAI's accelerating transformation into a for-profit company, while Microsoft continues to ramp up investment in its own proprietary AI capabilities. The rebalancing of their relationship reflects a new phase in the power dynamics between upstream and downstream players in the AI value chain.

Industry Outlook

From Xiaohongshu's AI governance practices to the restructuring of the Microsoft-OpenAI business relationship, we can observe the AI industry undergoing two critical transformations:

First, the accelerating development of governance frameworks. As AI content penetration continues to rise, platform-side rule-building has shifted from a "wait-and-see" approach to "proactive engagement." More content platforms are expected to follow suit with similar governance guidelines in the near future.

Second, deep adjustments to business models. The profit-sharing mechanisms between AI technology providers and distributors are being redefined, which will profoundly impact the evolution of the entire AI value chain.

Whether it concerns the healthy development of content ecosystems or the sustainability of commercial partnerships, the AI industry is transitioning from an era of "unchecked growth" to an "era of rules." For industry practitioners, understanding and adapting to these changes will be a critical and indispensable capability in future competition.