Yahoo Turned Its NBA Draft Expert Into an AI Chatbot. Brands Need the Rights Terms

Yahoo has put an AI version of a working sports analyst inside the product where his audience already goes for draft coverage. Ask Kevin O'Connor, launched for Yahoo Sports' NBA Draft guide, lets fans ask questions and receive AI-generated answers based on O'Connor's analysis, written voice, draft guide, coverage, show, and player or team data.

The product is small by design. Axios reported that Ask Kevin O'Connor is a limited-time NBA Draft feature, expected to stay live for several days after the draft before sunsetting by the end of June. Yahoo's official release frames it as one of two new Scout-powered answer experiences, alongside Ask Yahoo Scout in Yahoo Finance, and says the sports product gives fans answers written in O'Connor's style from his expert analysis.

The expert is the interface

The product's distinctive feature is its packaging. Yahoo did not create a fictional sports personality and ask users to pretend it had expertise. It took a named analyst, attached the interface to his annual draft guide, and made his existing work searchable through a conversational layer. Awful Announcing reported that the chatbot was trained on O'Connor's written work, podcast transcripts, hand-selected statistics, and scouting notes, and that O'Connor sees the feature as a complement to his normal articles and episodes.

That turns the creator asset from output into input. A draft guide, mock draft, podcast segment, scouting note, and byline normally travel as separate media units. In Ask Kevin O'Connor, they become source material for a single product that can answer a Warriors fan's player-fit question, redirect an unrelated query to Scout, and keep the user inside Yahoo's sports environment rather than sending them to a general-purpose assistant.

Scout is selling source control

Yahoo's broader Scout pitch is trust through visible sourcing and owned data. Business Insider reported that Scout launched in beta in January, is built on Anthropic's Claude and Microsoft's Grounding with Bing, and draws from the open web plus Yahoo's portfolio, including mail, news, finance, and publisher partners. Axios reported at launch that Scout uses Yahoo proprietary data, content, and insights alongside Claude and Bing grounding, with inline citations and source links meant to keep traffic moving back to publishers.

The O'Connor version applies that same logic to a person. The product needs enough of the analyst's material to feel specific, enough citations and grounding to avoid becoming a free-floating impersonation, and enough product boundaries to keep unrelated queries out of his name. The public descriptions show the source inputs and the intended limited-time scope. They do not show the prompt policy, answer-review workflow, error-correction process, or whether O'Connor can approve, reject, or audit individual answers.

The contract is the missing media model

For creator businesses, the unanswered part is commercial. O'Connor is a Yahoo Sports employee and senior analyst, which makes this cleaner than licensing an outside creator's likeness or archive. The public record still does not disclose whether the AI interface is covered by a specific talent agreement, whether O'Connor receives separate compensation, whether his approval is required for future Scout deployments, or how synthetic use of his style is bounded after the draft window closes.

Awful Announcing reported that Yahoo is exploring monetization for Scout experiences, including possible advertising, subscriptions, or a mix of open and paid features. That turns the talent question into a revenue question. If an expert's archive and voice make the interface useful, the next version has to answer whether the expert is paid like a writer, a host, a licensor, a product partner, or an employee whose output can be repackaged inside a new surface.

The creator-economy read

Most AI-creator coverage centers on synthetic influencers that imitate human presence without a real person behind them. Yahoo's model points to a more immediate media path: make an actual expert queryable, keep the response layer inside a publisher property, and use the expert's authority as the reason a user would choose that interface over a general answer engine.

That makes the diligence list more practical than philosophical. A brand, league, publisher, or creator-led company considering the same structure should know which assets train or ground the product, who owns the generated answers, what the user sees when an answer is AI-generated, how mistakes are corrected, whether the creator can pull material out, and how revenue is split if the interface starts selling ads, subscriptions, affiliate links, or sponsored placements.

Ask Kevin O'Connor is useful because the premise is narrow: one analyst, one draft guide, one sport, one time-bounded fan need. That is also the lesson. The first durable AI creator products may come from experts with deep archives and clear institutional homes, where the interface has a defined job and the source material is close enough to verify. The harder part is the contract layer, because the moment the interface becomes inventory, the creator's voice is no longer only editorial output. It is product infrastructure.