Video streaming platforms are accelerating their investments in artificial intelligence to reshape content creation and user engagement, with iQiyi emerging as a key player in this strategic shift. At the 2025 iQiyi World Conference held in Beijing on April 23, the company unveiled a comprehensive roadmap centered on AI-driven innovation across content production, recommendation systems, and interactive viewing experiences. This move reflects a broader industry trend where major Chinese streaming services are leveraging generative AI to meet evolving viewer demands for personalized, short-form, and highly engaging video content.
The conference, themed “With the Times, Dare to Lead, Create the Extraordinary,” featured presentations from iQiyi’s leadership, including founder and CEO Gong Yu, Chief Content Officer Wang Xiaohui, and President of Basic Architecture and Intelligent Distribution Liu Wenfeng. Over two days, executives detailed how AI is being integrated into every layer of the platform — from scriptwriting and video editing to real-time user behavior analysis and adaptive streaming. Wang Xiaohui emphasized that AI technology effectively addresses challenges in scaling personalized content delivery while maintaining creative quality, particularly in the rapidly growing micro-drama segment.
Central to iQiyi’s strategy is its “short, refined, abundant, and new” content approach, which prioritizes micro-dramas, shortened episode counts for long-form series, and concise films. According to Wang Xiaohui, “short” is the singular key word defining content strategy for 2025, responding to declining attention spans and rising demand for “simple satisfaction” alongside deeper thematic resonance. To support this, iQiyi has launched AI-powered tools such as the “Tiao Kan” (skip-watching) function and the “Peanut” personal AI assistant, designed to reconstruct viewing logic and enhance interactivity through intelligent scene navigation and personalized recommendations.
These innovations are part of a larger effort to diversify revenue beyond traditional subscriptions and advertising. During the conference, iQiyi officially announced its entry into the content e-commerce market, aiming to monetize viewer engagement through integrated shopping experiences tied to video content. This expansion into offline theme parks, lifestyle merchandise, and direct-to-consumer product offerings marks a significant evolution in the platform’s business model, aligning with industry-wide attempts to maximize lifetime user value.
AI as a Creative Force in Content Development
iQiyi’s investment in AI extends into the creative process itself, where generative models are being tested for script ideation, dialogue generation, and scene visualization. While specific technical partnerships were not disclosed at the conference, industry analysts note that Chinese tech giants like Baidu, Alibaba, and Tencent have similarly pursued AI-video integration, often leveraging proprietary large language models and diffusion techniques. IQiyi has previously collaborated with its parent company Baidu on AI initiatives, including the Ernie Bot model, which may inform its content generation tools.
Gong Yu described AI not as a replacement for human creativity but as an amplifier — enabling creators to experiment with narrative structures, reduce production timelines, and explore unconventional formats. He cited examples where AI-assisted editing reduced what once required three episodes to just thirty minutes of screen time, maintaining narrative impact while improving pacing. This efficiency gain allows iQiyi to increase output volume without compromising on artistic ambition, particularly in genres like fantasy, romance, and suspense where visual effects and complex plotting traditionally demand extensive resources.
Wang Xiaohui further clarified that the platform’s AI strategy balances “deconstruction and reconstruction” — using data insights to understand viewer preferences while preserving space for artistic innovation. By analyzing viewing patterns, drop-off points, and engagement spikes, AI helps identify which story elements resonate most, informing both algorithmic curation and creative decision-making. This dual focus aims to satisfy both the emotional escapism sought by audiences and their desire for narratives that explore life’s deeper questions.
Micro-Dramas and the Rise of AI-Optimized Formats
The spotlight on micro-dramas at the 2025 conference underscores their growing importance in iQiyi’s content slate. These short-form series, typically under 30 minutes per episode and often consisting of 10–20 installments, have gained traction among younger viewers seeking fast-paced, emotionally resonant stories. IQiyi has committed to expanding its micro-drama library, positioning the format as a key driver of future subscriber growth and engagement.
AI plays a supporting role in scaling this format by enabling rapid prototyping of plotlines, automated subtitle generation, and dynamic thumbnail optimization. Machine learning models assist predict which micro-drama concepts are likely to perform well based on genre trends, cast popularity, and thematic relevance to current events. This data-informed approach reduces creative risk while allowing room for experimental storytelling.
Importantly, iQiyi leadership acknowledged that micro-dramas are still maturing as an art form. Gong Yu expressed optimism that while current productions may lack the polish of traditional television, continued investment in talent, technology, and narrative depth will elevate the format over time. This long-term view reflects a belief that short-form content can evolve beyond mere snackable entertainment into a legitimate medium for meaningful storytelling.
Technological Infrastructure and User Experience Innovation
Beyond content creation, iQiyi is upgrading its technical infrastructure to support AI-intensive operations. Liu Wenfeng outlined advancements in intelligent content distribution, including adaptive bitrate streaming powered by real-time network and device analytics. These systems ensure smooth playback across varying bandwidth conditions while minimizing buffering — a critical factor in retaining user attention.
The “Peanut” AI assistant represents a step toward more conversational, context-aware interfaces. By understanding user preferences, viewing history, and even mood indicators (inferred from interaction patterns), Peanut can suggest content, explain plot points, or highlight thematic elements in real time. Such features aim to deepen user engagement by transforming passive viewing into an interactive experience.
Similarly, the “Tiao Kan” function allows viewers to skip repetitive or less engaging segments — such as prolonged exposition or predictable action sequences — based on AI-identified narrative beats. This functionality responds directly to viewer feedback about pacing issues in long-form content, offering a personalized viewing journey without requiring manual scrubbing.
Industry Implications and Competitive Landscape
iQiyi’s AI-forward strategy places it in direct competition with other major Chinese streaming platforms, including Tencent Video and Youku, which have also announced AI integration initiatives. Tencent has invested in AI-generated virtual idols and automated short-video creation tools, while Youku has explored AI-assisted dubbing and localization for international content. The race to implement generative AI responsibly and effectively is becoming a key differentiator in the saturated streaming market.
Globally, the trend mirrors developments at Western platforms like Netflix and Disney+, which have experimented with AI for content recommendation, localization, and even preliminary script analysis. However, Chinese platforms appear to be moving more aggressively toward end-to-end AI integration in production workflows, possibly due to stronger domestic AI research ecosystems and less restrictive data governance environments.
Analysts caution that while AI offers significant efficiency gains, challenges remain regarding copyright, ethical use of training data, and the potential homogenization of creative output. IQiyi has not publicly detailed its AI training sources or governance frameworks, leaving questions about data provenance and model transparency unresolved. As regulatory scrutiny increases worldwide, platforms will necessitate to balance innovation with accountability.
What’s Next for iQiyi and AI in Streaming
The company has not announced a specific timeline for rolling out all AI features unveiled at the conference, but indications suggest a phased implementation throughout 2025 and into 2026. Viewers can expect to see gradual enhancements in recommendation accuracy, interactive viewing options, and micro-drama availability as these systems mature.
For industry watchers, the next key developments to monitor include any official technical disclosures from iQiyi regarding its AI models, partnerships with AI research institutions, or pilot programs involving generative content. User feedback on features like Peanut and Tiao Kan will be critical in assessing their real-world impact on satisfaction and retention.
As the streaming landscape continues to evolve, iQiyi’s bet on AI reflects a broader conviction that the future of video entertainment lies not just in what we watch, but in how we discover, interact with, and derive meaning from it. Whether this technological shift will deepen audience connection or introduce new complexities remains to be seen — but for now, the platform is positioning itself at the forefront of the industry’s next transformation.
For updates on iQiyi’s AI initiatives and corporate announcements, readers can refer to the company’s official investor relations portal and press release archive.
If you found this analysis informative, consider sharing it with others interested in the intersection of technology and entertainment. Join the conversation by leaving a comment below — what role do you feel AI should play in shaping the stories we love?