Elon Musk's xAI, Meta And Google Sued By New York Times Reporter John Carreyrou Over Alleged Use Of Pirated Books To Train AI

Investigative reporter John Carreyrou and five other authors filed a federal lawsuit in California accusing OpenAI, Google, Meta, xAI, Anthropic and Perplexity of using copyrighted books without permission to train large language models, deliberately opting out of a class-action route to preserve individual remedies. The complaint highlights Anthropic's prior $1.5 billion class settlement and comes as these AI firms carry very large private/public valuations (OpenAI fundraising talks up to $100B potentially valuing it as high as ~$830B; Perplexity raised $200M at a $20B valuation; xAI and Anthropic have multi‑billion funding/IPO plans), creating meaningful liability and licensing risk that could affect future capital raises, valuations and operating costs for AI companies.

Analysis

Market structure: Litigation raises direct cost for LLM training providers (OpenAI/Alphabet/Meta/xAI/Anthropic/Perplexity), increasing marginal cost of data and licensing. A conservative stress: $1–5B of legal/settlement reserve per large defendant would shave ~50–200 bps off operating margins for Google/Meta-equivalent AI segments, pressuring near-term EPS and capex for model training. Winners include cloud/data licensors, publishers and incumbents with proprietary datasets; losers are uncapitalized AI pure-plays and secondary-market valuations of private AI rounds.

Risk assessment: Tail risk includes injunctions halting model retraining, statutory damages up to $150k per work, or a precedent setting $1B+ settlement (Anthropic was $1.5B). Near-term (days–weeks) expect volatility spikes and IV increases; short-term (3–9 months) risk centers on settlements and fundraises; long-term (1–3 years) regulatory licensing markets and compulsory licensing could normalize costs. Hidden dependency: enterprise AI purchasing may accelerate shift to licensed, auditable models, raising switching costs and favoring deep-pocketed incumbents.

Trade implications: Tactical alpha lies in volatility and relative exposure. Expect credit spreads on smaller AI/private financings to widen 100–300 bps; equity-wise, defensive large-cap tech with diversified revenue (AAPL, MSFT) will outperform concentrated AI-data exposed names. Options strategies that monetize higher IV (buy puts on GOOG/GOOGL, sell premium against long positions) are attractive over 3–6 month windows while awaiting rulings.

Contrarian angle: Market may overprice existential threat — even $10B aggregate settlements are <0.5% of combined market caps of Alphabet/Meta and manageable via amortized licensing. Historical parallel: music sampling litigation created licensing markets and recurring revenue for rights holders. Unintended consequence: a paid, audited dataset market emerges, benefiting data brokers, cloud vendors and companies that control proprietary user data.

Key Decisions for Investors

• Establish a 1.5% NAV hedge short on GOOGL via 3-month 5% OTM put purchases (rollable) — target a 10–20% downside or IV-driven 50–100% premium capture; cut if shares trade above entry by +5% for 2 consecutive sessions or if a favorable court opinion is issued for defendants within 90 days.
• Initiate a 2% NAV pair: long AAPL (buy shares) and short META (equal dollar notional) to express divergence — AAPL has lower legal/data exposure; exit if AAPL/META relative outperformance reverses by >7% over 60 days or after a definitive regulatory settlement is announced.
• Reduce high-risk private/public small-cap AI exposure by 30% immediately and redeploy into cloud/data infrastructure names (e.g., GOOG cloud revenue exposure, MSFT) or data-licensing plays; target redeployment over 30 days and reassess after major filings/settlements.
• Trade volatility: buy 3–6 month straddles or puts on select AI-exposed names (GOOG/GOOGL/META) sized at 0.5–1% NAV where implied volatility is <historical realized +50%; take profits on IV spikes >40% from entry or after material legal milestones (filing, settlement, injunction) within 90 days.