JioHotstar and OpenAI are rolling out a ChatGPT-branded conversational layer that lets nearly half a billion monthly users search by voice or text, access live sports insights, and surface the streamer’s 300,000-hour, 19-language catalogue within ChatGPT
JioHotstar has teamed up with OpenAI to bring a ChatGPT-branded conversational assistant into its streaming app. Rather than scrolling through endless menus or exacting keyword searches, viewers can now speak or type natural requests and get context-aware, culturally tuned recommendations—everything from what to watch with family tonight to highlights from a live match.
How it works
The assistant uses multilingual cognitive search to read intent from ordinary language. Instead of matching literal keywords, it picks up situational clues—a description of a mellow weekend dinner, someone saying they want “nostalgic comedies,” or a request in a regional dialect—and surfaces shows, clips or live moments that fit. You can interact by voice or text, and the system supports both instant, live queries and on-demand recommendations for later browsing.
Across live and on-demand content
This feature isn’t limited to on-demand libraries. For live events—especially sports—it can pull up match stats, identify key moments, and answer player-specific questions in near real time, while keeping the primary broadcast front and center. For on-demand viewing, the assistant draws from JioHotstar’s catalog of more than 300,000 hours across 19 languages and over 100 channels, spanning originals and licensed content. Behind the scenes, conversational cues are mapped to catalog metadata and editorial signals so the suggestions feel both relevant and immediate.
Multilingual, flexible, and context-aware
A core aim is removing language friction. The assistant handles code-switching and regional expressions, so a mid-query switch from Hindi to Tamil (or to English) won’t break the flow. That makes discovery more inclusive for India’s diverse audience and helps surface long-tail or regional titles that traditional search might miss.
Two-way discovery with ChatGPT
This partnership isn’t one-sided. JioHotstar’s catalogue will be discoverable from within ChatGPT, meaning users can ask the chatbot for viewing suggestions and receive direct links back to the streaming service. Conversely, recommendations found inside the JioHotstar app can point users toward deeper context or follow-up conversations through the chatbot. The result: discovery can start inside the app or begin in a chat and end with playback.
Design and product strategy
The product is built around low-friction inputs, fast retrieval for near-real-time queries, and suggestions that respect the viewing context—keeping live coverage uninterrupted while layering helpful, interactive information on top. Initial launches will expose a limited feature set and expand gradually, allowing the team to refine responses based on real user behavior and to phase in additional formats over time.
Business and privacy considerations
Analysts expect the assistant to shorten the path from discovery to playback, which could boost engagement and surface more regional and niche titles—good news for rights holders and monetisation. At the same time, exposing recommendations through an external chatbot could change traffic and ad patterns, and potentially extend reach beyond registered users.
Those gains come with trade-offs. Data sharing between a streaming platform and a third-party chatbot raises questions about consent, profiling and how promotional content is surfaced. JioHotstar and OpenAI will need clear user controls, transparent data flows, and solid safeguards as features roll out.
Where this fits in a larger strategy
The move puts JioHotstar among the first Indian streamers to stitch conversational AI directly into the viewer journey. The company’s parent, JioStar (formed after the Disney Star–Viacom18 merger), frames the assistant as part of a broader AI push—everything from creative ideation and production planning to targeted distribution and monetisation. Executives say the technology can encourage more participatory viewing while making it easier to navigate huge, multilingual catalogs.
Practical examples
Imagine hosting relatives with different tastes: instead of toggling filters, you describe the group and get a short, curated list everyone can choose from. Or picture two viewers hunting for niche films about twin protagonists—the assistant returns options matched for tone, era and availability. During a live game, ask for “turning points in the third quarter,” and the system locates clips, stats and quick summaries far faster than manual scrubbing.
How it works
The assistant uses multilingual cognitive search to read intent from ordinary language. Instead of matching literal keywords, it picks up situational clues—a description of a mellow weekend dinner, someone saying they want “nostalgic comedies,” or a request in a regional dialect—and surfaces shows, clips or live moments that fit. You can interact by voice or text, and the system supports both instant, live queries and on-demand recommendations for later browsing.0