According to the CEO of Level-5, 80 to 90% of games are created using AI

Level-5's CEO stated that roughly 80–90% of games are created using AI, signalling broad adoption of AI-driven tools in game development for asset creation and design. The claim implies potential cost and time efficiencies for studios and increased addressable market for AI tool vendors, but contains no company financials or verifiable metrics, so investors should treat it as a directional industry comment rather than a direct earnings or revenue signal.

Analysis

Market structure: If 80–90% of games are being created with AI, winners are AI-infrastructure and engine providers (NVDA, MSFT, GOOGL, AMZN, U, ADBE) that capture compute, tooling and distribution margins; losers are labor-heavy mid/small devs, art outsourcers and QA shops as per-title development cost could fall 30–60% over 12–24 months, increasing title supply but compressing pricing power for undifferentiated games. Competitive dynamics favor platforms with discovery and network effects (Roblox/RBLX-like platforms, Steam/Valve private) which can monetize attention as content supply explodes, shifting market share to a smaller set of aggregators.

Risk assessment: Tail risks include rapid regulatory/IP rulings (major copyright losses >$500m) or export controls on compute that spike GPU prices 20–50% and derail margins; immediate market moves should be muted, but expect measurable KPI inflection in 1–3 quarters as studios report lower unit dev costs and mixed engagement. Hidden dependencies: datasets/licenses, cloud compute pricing and content moderation costs — second-order effects may increase opex for publishers despite lower dev capex. Catalysts: Unity/Unreal AI integrations, NVDA earnings beats, and any EU/US AI training legislation within the next 30–90 days.

Trade implications: Overweight semiconductor/AI infra and cloud (NVDA 2–3% position, MSFT 1–2%, AMZN/GOOGL 1% each) and underweight mid-cap gaming/gaming ETF exposure (reduce GAMR exposure by 50%); implement pair trades (long NVDA vs short GAMR) to isolate AI compute upside vs content dilution. Use options to express convexity: 6–12 month NVDA LEAP call or 3–6 month put spreads on GAMR to hedge consumer attention risks; act within 2–6 weeks to capture product rollouts and holiday selling season data.

Contrarian angles: Consensus underestimates discovery friction — more titles do not equal revenue; short-term investor euphoria for “AI-made” games may be overdone, creating mispricings in small-cap devs whose revenue per title could decline 20–40% over 12–18 months. Historical parallels: mobile app gold-rush followed by consolidation — expect similar winner-take-most outcome, and an unintended consequence is that costly litigation or licensing could re-concentrate economic rents to large incumbents, making large-cap infra names safer havens than growthy small devs.

Key Decisions for Investors

• Establish a 2–3% long position in NVIDIA (NVDA) within 2 weeks, target +30–60% over 12 months on sustained AI compute demand; set a 15% stop-loss and trim to 1% if NVDA misses two consecutive quarters of datacenter growth.
• Add 1–2% long exposure to Unity Software (U) or Adobe (ADBE) within 1 month to play engine/tooling monetization (horizon 6–18 months); trim if MAU/engine revenue growth lags by >200 bps vs consensus in next two earnings cycles.
• Cut direct exposure to small/mid-cap game developers and the VanEck Video Gaming & eSports ETF (GAMR) by ~50% over 30 days; initiate a 1–2% notional 3–6 month put spread on GAMR (buy 20% OTM puts, sell 40% OTM puts) to hedge dilution/attention risk.
• Monitor regulatory/IP developments (US Copyright Office, EU AI Act, major class-action suits) on a 30–90 day cadence and commit to reducing AI-tooling or platform longs to <1% if an adverse ruling implies >$500m in industry-wide licensing/fine exposure within that window.