S16.9 · Travel, Hospitality & Leisure
Wagering venues and betting operators, split between licensing-gated casino property and asset-light online betting.
Global gross gaming revenue is estimated at $700-800B (H2 Gambling Capital and Statista-sourced industry figures, 2024-2025) across land-based, online and lottery formats, though figures vary materially by source and scope; online gambling alone is sized separately at $100-115B (Statista, Grand View Research, 2025). Treat this as two segments wearing one label: capital-intensive, licensing-gated land-based casino property on one side, an asset-light, fast-consolidating online-betting business on the other. AI reaches them very differently, and mostly reaches the second.
Online and mobile betting grows roughly 10-12% annually; land-based casino growth is low-single-digit. US commercial and tribal gaming and Macau lead land-based revenue; the UK, continental Europe and the post-PASPA US market lead regulated online betting. Land-based gaming is moderately concentrated among Las Vegas Sands, MGM Resorts, Caesars Entertainment, Wynn Resorts and Galaxy Entertainment; online betting concentrates among Flutter/FanDuel, DraftKings, Entain and Bet365. Upstream, gaming-equipment and software suppliers including IGT, Light & Wonder and Evolution serve both channels; downstream, state and national regulators and marketing and affiliate channels govern how operators reach customers.
Casino real estate is a highly capital-intensive, licensing-gated asset class, commonly separated from casino operations through sale-leaseback structures, with specialist real estate investment trusts such as VICI Properties and Gaming and Leisure Properties holding the underlying property. Online betting is the opposite — asset-light and scalable, though customer-acquisition cost runs high — and nearly all of the segment's recent consolidation activity has concentrated there.
Odds-setting and trading-desk labor is where the automation lands first, particularly at smaller sportsbooks that lack the scale to build proprietary pricing models: AI-driven pricing is increasingly available as a service, lowering the cost of running a book. What that commoditizes is the proprietary trading model itself. If smaller operators can price competitively without in-house quantitative teams, the historical differentiation of the larger books narrows. But keep the moats in order. State-by-state licensing, not technology, remains the primary and far more durable moat in this segment — it gates market entry regardless of how sophisticated an operator's pricing engine is, and no pricing engine substitutes for a license.
Media companies licensing their brands into sports betting is expansionary for both sides: media reach extends into a wagering product, and betting operators gain an established audience. The trading-automation effect plays out over 2-5 years. The licensing moat persists on a much longer horizon, largely indifferent to AI.