IBM launched limb tracking at this year’s U.S. Open – and it’s scoring every tennis player’s serve for fans to see

IBM and the USTA launched an AI-powered "serve quality" feature in the 2026 U.S. Open app, scoring players' serves out of 100 using camera-based tracking of 21 limbs and joints. IBM expects to analyze 1.2 billion joints and generate 7 million serve-quality insights during the tournament. The feature, built with WatsonX and Hawk-Eye data, could expand to other tennis strokes, Grand Slams including Wimbledon, and sports such as golf.

Analysis

This is strategically useful for IBM’s consulting/software narrative but immaterial to near-term financial estimates. The value is not the fan-facing score itself; it is proof that watsonx can sit atop third-party computer-vision infrastructure, convert unstructured motion data into an intelligible product, and monetize an existing sports-rights relationship without requiring IBM to own the capture hardware. That lowers the deployment hurdle for adjacent venue, broadcast, insurance, physical-therapy, and training applications.

The key 1-3 month catalyst is whether IBM discloses engagement metrics, renewal economics, or additional deployments around its sports partnerships. A Wimbledon or Masters expansion would validate reusability, but investors should not extrapolate consumer-app usage into material AI revenue: sports activations have historically been brand and client-acquisition tools, with commercial conversion opaque. The more consequential read-through is for Sony Group (SONY), whose Hawk-Eye ecosystem gains another data layer and can potentially sell upgraded tracking workflows to rights holders.

Contrarian view: IBM shares may receive a modest AI-productivity halo, but this does not resolve the market’s central underwriting question—whether generative-AI bookings convert into durable, high-margin software revenue rather than labor-intensive consulting. The feature’s post-match availability also limits live betting and media-rights monetization, the highest-value use cases; a transition to reliable real-time outputs would be the meaningful inflection. Falsify the cautious view if IBM begins quantifying incremental software ARR or signs multi-event contracts that use the same stack outside its sponsored properties.

Key Decisions for Investors

• No standalone directional trade in IBM on this launch; treat it as a qualitative validation only. Reassess after the next earnings call if management identifies AI software-bookings growth, watsonx recurring revenue, or conversion from sports/brand engagements into enterprise contracts.
• Maintain a 6-12 month watchlist long SONY versus a broad media/entertainment proxy (e.g., short XLC) if Hawk-Eye announces paid, multi-tournament skeletal-tracking deployments. The thesis requires evidence that venue operators—not sponsors—are funding hardware/software upgrades.
• For existing IBM longs, use any AI-led rally without corresponding bookings or free-cash-flow guidance improvement to reduce exposure. A stronger signal would be a widening gap between AI commentary and software segment growth; upside invalidation is disclosed recurring revenue or a material expansion in software gross margin.
• Monitor real-time permissions and data-rights developments at major sports leagues over the next 6-18 months. If tracking outputs become live and licensable, sports-data vendors and betting-data infrastructure names could capture more economics than IBM, which currently appears positioned primarily as the analytics layer.