On July 28, 2026, HBO Max became the last of the major streamers to ship AI-driven discovery, and it did it with two features at once. The first, HBO Max Shorts, is a personalized vertical feed of clips, trailers, and bonus content — the TikTok grammar now familiar from Netflix Clips and Disney+ Verts — but the interesting part is how it is built: an in-house AI analyzes scene-level metadata across thousands of hours of film and TV to find the moments most likely to hook a viewer, then human editors pick the final clips and cut them to vertical. The second, Ask HBO Max, is a conversational search that takes plain-language prompts like "in the mood for a comedy" or "dysfunctional family drama" and returns titles by understanding intent, using semantic search that matches meaning even when your words never appear in a show's metadata. Read together, the two features are not really about short video — they are about a shift in how catalogs get discovered: from titles and keywords to machine understanding of what a piece of content is about and what a viewer actually wants. This guide explains exactly what HBO Max shipped and on what dates, how the scene-level and semantic AI works, why the whole industry moved to intent-based discovery at once, and — the part that matters if you make content for a living — what an AI-driven, intent-matching discovery layer changes about how you should produce and shape your own short-form, and the honest limit that these streaming feeds are closed gardens you cannot upload to.
On July 28, 2026, HBO Max shipped two features at once, and while the headlines called them "short video and AI search," what they actually mark is a change in how a streaming catalog gets discovered. The first, HBO Max Shorts, is a personalized vertical feed of clips, trailers, and bonus content — the TikTok grammar already familiar from Netflix Clips and Disney+ Verts. The second, Ask HBO Max, is a conversational search that takes plain-language prompts like "in the mood for a comedy" and returns titles by understanding what you mean rather than matching what you typed. Warner Bros. Discovery is the last of the major streamers to add both, which is precisely what makes the moment worth reading: the whole industry has now converged on AI-driven, intent-based discovery.
This guide covers what HBO Max actually launched and on what terms, how the two AI systems work — scene-level clip selection and semantic search — why every major platform arrived at the same place at once, and the part that matters if you produce content: what an intent-matching, AI-driven discovery layer changes about how you should shape your own short-form, and where it changes nothing. It is deliberately distinct from its nearest neighbor. The full field of streaming feeds — Netflix, Disney+, Peacock, Tubi — and the format-convergence argument live in streaming platforms are pivoting to creator-style short-form; this page is about the discovery half specifically: the AI that decides which moment to surface and which title to recommend, and the intent-over-keywords shift underneath it.
Two separate features, launched together, aimed at the same problem — helping subscribers find something to watch in a catalog too large to browse. It is worth pinning down each precisely, because the details are where the real trend lives.
Shorts is a vertically scrolling feed of trailers, clips, and bonus content drawn from the HBO Max catalog and personalized to each viewer based on watch history. You can start the full title immediately from a clip or save it to a watch-later list — so the feed is explicitly a discovery layer that funnels you into long-form, not a replacement for it. At launch it is a limited test: available to select iOS users in the US, with HBO Max saying it will expand to more devices, international markets, and sports content over time. Functionally it is the same move Netflix made with Clips and Disney+ with Verts — a lean-forward, feed-first surface bolted onto a lean-back streaming app.
The second feature is the more telling one. Ask HBO Max is a conversational search, tested first with select Android users in the US, that accepts natural-language prompts — the examples HBO Max gave include "in the mood for a comedy," "dysfunctional family drama," and "best movie for a girls night in" — and recommends titles by interpreting the intent behind the phrase. It combines natural-language understanding with semantic search, which means it can match a request to a show even when the exact words you used appear nowhere in that show's metadata. That is the difference between searching a database and being understood: you no longer have to know the title, the genre tag, or the right keyword. You describe the mood, and the system maps meaning to catalog.
The two features run on two different but related pieces of machine understanding, and both are worth understanding because the same techniques govern how content gets found on the open platforms creators actually use.
Shorts is built on scene-level metadata. Rather than treating a two-hour film as one opaque object, the in-house AI analyzes the content at the level of individual scenes across thousands of hours of film and television, tagging the moments most likely to engage a viewer — the hook, the laugh, the reveal, the cliffhanger. That is what lets the system pull a compelling thirty-second vertical moment out of a buried catalog title that a browse grid would never surface. Crucially, it is not fully automated: HBO Max editors then decide which of the AI-surfaced clips best represent the library and cut them to vertical. The AI provides scale — finding candidate moments no human could review by hand — and humans provide taste. That human-in-the-loop split is a pattern worth noting, because it is the honest version of "AI-powered" that most durable systems actually use.
Ask HBO Max runs on semantic search — representing both the user's request and every title by meaning rather than by literal words, then matching them in that meaning-space. This is the same shift that has been rewriting web search, where engines increasingly answer the intent behind a query instead of ranking pages by keyword overlap, a change covered in SEO in the age of AI search and AI search behavior replacing keywords. A streaming catalog is just another corpus to be searched by intent, and HBO Max applying it to "what should I watch tonight" is the same technique pointed at a different index. The takeaway for anyone who publishes content: discovery, everywhere, is moving from string-matching to meaning-matching.
The reason HBO Max being "last" matters is that it completes a set. Netflix shipped its vertical Clips feed in spring 2026 and pairs it with its own AI-powered search; Disney+ rolled out a vertical feed called Verts; Peacock and Tubi added short-form feeds earlier. With HBO Max joining in July, every major US streaming platform now runs an AI-driven, feed-first discovery layer on top of its catalog, and several pair it with intent-based search. That is not a coincidence of product roadmaps. It is the whole category concluding, independently and simultaneously, that the old discovery model — a grid of box art you scroll past — no longer works for catalogs this large and attention this short.
And the pattern is not confined to streaming. The same two moves — a personalized vertical feed and an intent-understanding search — are what already define discovery on TikTok, YouTube, and Instagram, and increasingly on the web via AI answer engines. When the same shape appears in social apps, streaming apps, and search at once, it stops being a trend inside any one industry and becomes the ambient condition every publisher operates in: your content is found because a machine understood what it is about and matched it to what someone wanted, not because someone typed your exact title. The consequences of that for reach are worked through in the publisher traffic collapse and AI discovery.
Two pressures forced it. The first is catalog size against attention span. A modern streaming library holds tens of thousands of titles, and the classic "decision paralysis" problem — subscribers spending more time browsing than watching, or bouncing to a social feed instead — is a direct threat to engagement and, therefore, to retention. An AI that can surface a compelling thirty-second moment from a title you would never have clicked, or answer "something funny and light" without making you guess a name, is a direct attack on that paralysis. The second pressure is competitive: attention has concentrated in short-form, feed-first video, and streamers watched YouTube and TikTok pull away on daily engagement. Copying the discovery grammar that won is the defensive response.
There is a deeper reason the timing lined up, though: the underlying AI got good enough and cheap enough at the same time. Scene-level analysis across thousands of hours of video and reliable semantic search over a large catalog were expensive research problems a few years ago and are commodity capabilities now. When a technique becomes affordable, every well-funded player in a category adopts it in the same window — which is exactly why HBO Max, Netflix, and Disney+ all arrived at the same feature set within a few months of each other.
Start with what it does not change, because this is where the trend is easiest to misread. HBO Max Shorts is not a place you can publish. Like Netflix Clips and Disney+ Verts, the feed is stocked entirely from the platform's own catalog — AI-surfaced clips of its own films and shows, chosen by its own editors. There is no upload button, and "get my content onto HBO Max" is the wrong goal. Reading this launch as a new distribution channel is the same mistake people made with Netflix Clips, and it points effort at a door that is closed.
What it does change is the environment your content already lives in. Two signals matter. First, format: when the last holdout streamer rebuilds discovery around vertical, feed-first, personality-led clips, it removes any remaining doubt that short-form vertical is the universal container for video — a case made in full in short-form video on mobile is the default now. Second, and more specific to this launch: discoverability is now a function of machine understanding. The systems deciding who sees your work — social feed algorithms, YouTube and TikTok search, AI answer engines — increasingly surface content by understanding what it is about and matching it to a viewer's intent, exactly like Ask HBO Max maps "in the mood for a comedy" to a catalog. That raises the value of producing short-form that is legible to those systems and that answers real intents, not just content that exists.
If discovery is shifting from keyword-matching to intent-matching, the practical creator response is to stop making one generic post per idea and start making several, each shaped to a distinct intent someone might actually express. Ask HBO Max is a useful model here: it does not need a viewer to know a title; it needs the catalog to contain something that satisfies "a dysfunctional family drama." The parallel for a creator is that a single idea can answer many intents — the how-to version, the mistake-to-avoid version, the beginner's version, the quick-win version — and each one is a separate hook that an intent-based discovery system can match to a different searcher or scroller. One idea, many intent-shaped angles, is the production posture an AI discovery layer rewards.
The other half of the lesson is that clips are how catalogs get discovered — and your back catalog is a catalog too. HBO Max's scene-level AI exists to pull the compelling moment out of a long title so it can travel as a short. A creator with a library of long-form videos, streams, or podcasts sits on exactly the same opportunity: the strongest moments, cut to vertical, are the discovery surface for the full piece. The mechanics of doing that well are in short-form AI clips from long-form content, and the underlying term is defined in short-form video. The streamers are turning their catalogs into feed-native clips because that is how anything gets found now; the same move, pointed at the open platforms, is available to you.
Keep three boundaries in view. This is an early, limited test — Shorts on select iOS users, the search on select Android users, both US-only, with expansion promised but not delivered — so the specifics will shift and anything written about the exact rollout has a short shelf life. The scene-level AI is not magic autonomy either; it surfaces candidates and humans still choose and cut, which is a feature, not a shortcoming, but it means "AI-powered discovery" is really "AI-accelerated, human-curated discovery," and the human taste is doing real work. And most importantly for a creator, none of it is a channel you can push into. The value is entirely in the signal: discovery has gone AI-driven and intent-based across every surface at once, and the useful response is to produce accordingly for the platforms that are open, not to wait for a streaming upload button that is not coming.
The discovery shift this launch confirms has a clear implication, and it is a production one. If machines now surface content by understanding it and matching it to intent, then the winning move is to produce more intent-shaped angles of your ideas, as short-form the discovery systems can read, at a cadence a manual workflow cannot hold. That is a throughput problem, and it is the exact gap Kompozy is built to close. The honest boundary first: Kompozy does not get you into HBO Max Shorts or any streamer's feed — those are closed catalogs. It is the engine for the open surfaces, where your uploads actually land and where the same AI-driven, intent-based discovery decides who sees them.
The specific leverage is intent variety from one idea. Where the discovery layer wants many angles — the how-to, the mistake, the quick-win, the beginner take, each a different hook for a different searcher — Kompozy generates them in one pass from a single brief rather than one post at a time. It is a generation engine, not a repurposing shim, so it builds net-new content for each angle: a Persona Shorts talking-head clip with a consistent face and burned-in captions for one intent, a Carousel Post for another, Quote Graphics and text posts for the intents that read better as stills, plus blogs and newsletters for the surfaces search reaches. And Clipped Shorts do to your long-form library exactly what HBO Max's scene-level AI does to its catalog — pull the strongest moments out as vertical, feed-native clips — except aimed at the platforms open to you. The repurposing discipline underneath it is defined in content repurposing.
What turns that from raw output into a discoverable, on-brand body of work is consistency and cadence. A Persona Brief governs the voice so every angle reads as one identity instead of a scattered set of guesses, HyperFrames render the graphics pixel-exact to your brand, and Autopilot with a per-post review gate keeps the whole spread shipping across eight social platforms plus blog and email on schedule. So while the streamers rebuild their apps around AI discovery you cannot upload to, the useful move is to feed the discovery layers that are open — with more intent-matched short-form than you could make by hand, produced from one brief and shipped everywhere it counts. That is the same owned-format discipline behind short-form content strategy in 2026: the format and the discovery model everyone now agrees on, produced on-brand and at a volume the shift just made more valuable.
HBO Max's July 28, 2026 launch of Shorts and Ask HBO Max completes an industry-wide move: every major streamer now runs an AI-driven, feed-first, intent-based discovery layer, using scene-level clip selection and semantic search to match a huge catalog to a viewer who no longer wants to browse. The features are not an upload channel for creators — the feeds are closed gardens filled from the platform's own library. The signal is what counts: discovery everywhere has shifted from titles and keywords to machine understanding of content and intent, which rewards producing many intent-shaped short-form angles that AI systems can read and surface. The streamers are turning their catalogs into discoverable clips because that is how anything gets found now — and the same move, produced at volume for the open platforms, is the one available to you.
They are two AI-powered discovery features Warner Bros. Discovery launched for HBO Max on July 28, 2026. HBO Max Shorts is a personalized, vertically scrolling feed of clips, trailers, and bonus content pulled from the catalog and tailored to your watch history. Ask HBO Max is a conversational search that takes natural-language prompts — "in the mood for a comedy," "dysfunctional family drama" — and recommends titles by interpreting intent rather than matching exact keywords. Both are early tests: Shorts on select iOS users in the US, the search on select Android users in the US.
An in-house AI tool analyzes scene-level metadata across thousands of hours of film and TV to identify the moments most likely to engage a viewer — the beats a human might tag as a hook, a laugh, or a cliffhanger. It is not fully automated, though: HBO Max editors then decide which of those AI-surfaced clips best represent the library and adapt them to the vertical format. So the AI does the scale (finding candidate moments in a huge catalog) and humans do the taste (choosing and cutting the final clips).
Normal search matches your typed words against titles, cast, and genre tags, so it fails when you do not know what you want. Ask HBO Max uses natural-language understanding plus semantic search to interpret intent — it maps a phrase like "best movie for a girls night in" to the meaning behind it and returns relevant titles even when those exact words appear nowhere in a show's metadata. It is the same intent-over-keywords shift reshaping web search, applied to a streaming catalog.
No — it is the last major one to add it, which is the real signal. Netflix launched its vertical "Clips" feed in spring 2026 and has its own AI-powered search; Disney+ rolled out a vertical feed called "Verts"; Peacock and Tubi added short-form feeds earlier. HBO Max joining completes the set: every major streaming platform now runs an AI-driven, feed-first, intent-based discovery layer on top of its catalog. The convergence is the story, not any single feature.
No. Like Netflix Clips and Disney+ Verts, the HBO Max Shorts feed is stocked entirely from the platform's own catalog — AI-surfaced clips of its films and shows, chosen by its editors — not an open upload button. So this is not a new distribution channel for independent creators. The value for creators is the signal, not the pipe: it confirms that discovery everywhere is moving to AI-driven, intent-matched, vertical feeds, which raises the payoff of producing short-form that a machine can understand and surface on the platforms that are actually open to you.
HBO Max launched two AI-powered discovery features on July 28, 2026: HBO Max Shorts, a personalized vertical feed of clips, trailers, and bonus content, and Ask HBO Max, a conversational search that answers natural-language prompts like "in the mood for a comedy." Shorts uses an in-house AI analyzing scene-level metadata across thousands of hours of content, with human editors picking the final vertical clips. It makes HBO Max the last major streamer to add AI discovery, following Netflix Clips and Disney+ Verts — confirming that discovery everywhere is shifting from titles and keywords to machine understanding of content and intent.
Get started → · ← All guides · Compare Kompozy vs other tools