Decoding Two AI Lawsuits: xAI vs Apple/OpenAI and Nikkei/Asahi vs Perplexity — Issues, Legal Battlegrounds, and Industry Impact 【as of Aug 26, 2025】

This article examines two lawsuits in detail from the perspectives of “What is at stake,” “Key legal points for victory,” and “Impacts on products, monetization, and regulation”:
① xAI (Elon Musk) suing Apple and OpenAI in the U.S. federal court in Texas, and
② Nikkei and Asahi Shimbun suing Perplexity in Tokyo District Court.

TL;DR (Bottom Line First)

• xAI’s U.S. antitrust suit challenges iOS’s default integration of ChatGPT and the reduced visibility of its own Grok, framed as monopolization and collusion. The crux will be market definition, proof of exclusionary conduct, and consumer welfare. Apple/OpenAI are expected to counter with arguments about integration benefits, non-exclusivity, and alternative access routes.
• Nikkei/Asahi’s copyright suit alleges unauthorized copying, storing, and reuse by Perplexity, claiming infringement of reproduction rights and making available rights. The focus will be on the scope of Japan’s TDM (text and data mining) exception and how outputs are treated. Plaintiffs seek ¥2.2 billion each in damages plus injunctive relief.

1) xAI (Musk) vs Apple/OpenAI: Is “Default Integration” and App Visibility an Antitrust Violation?

Case Overview

• Court: U.S. Federal Court, Texas
• Defendants: Apple, OpenAI
• Plaintiffs: xAI and X (formerly Twitter)
• Claim: Apple deeply integrated ChatGPT into iOS and gave it preferential exposure and rankings on the App Store, allegedly in collusion with OpenAI, thereby disadvantaging xAI’s Grok. Plaintiffs allege anticompetitive exclusion/monopolization and seek damages and injunctive relief.

xAI’s Legal Storyline (Expected Claims)

• Sherman Act §1 (collusion) / §2 (monopolization):
  • Market definition: Narrow framing such as “AI assistant distribution on mobile OSs” or “access points to generative AI within iOS.”
  • Exclusionary conduct: Default settings + deep OS integration + self-preferencing (exposure, ranking) framed as exclusion.
  • Consumer harm: Fewer choices, reduced innovation, worsened pricing/terms.
• Relief sought: Injunction (forcing changes/opening access) + substantial damages.

Apple/OpenAI’s Likely Counterarguments

• Market is broad: Alternatives exist via the web, other OSs, and apps—competition not foreclosed.
• Integration justified: Improves performance, safety, UX — a pro-competitive efficiency. Not exclusive; others can integrate.
• Exposure rules: App Store rankings are based on quality/safety criteria, not collusion.

Keys to Outcome

1. Market definition: Narrow (iOS-only AI access) → favors xAI; broad (all AI access) → favors Apple/OpenAI.
2. Proof of exclusion: Internal docs showing deliberate exposure manipulation = strong for xAI. Otherwise, UX improvements are defensible.
3. Consumer welfare: Quantitative proof of reduced price/quality/choice will be pivotal.

Industry & Monetization Impact

• Intensifies the battle between native (on-device) AI vs external apps.
• Reignites scrutiny of “default” power under antitrust law.
• CAC (customer acquisition costs) must be reassessed in light of OS/gatekeeper regulation. Settlement or regulation could push for transparency in exposure rules.

2) Nikkei & Asahi vs Perplexity: “Unauthorized Copying, Storage, and Reuse” — What’s at Issue under Japanese Law

Case Overview

• Court: Tokyo District Court
• Plaintiffs: Nikkei Inc., Asahi Shimbun Co.
• Defendant: Perplexity AI (U.S.)
• Claims: ¥2.2 billion each in damages + injunctive relief (possibly including preliminary injunction). Complaint alleges unauthorized copying, storage, and use of news content.

Rights at Stake under Japanese Copyright Law

• Reproduction right (Art. 21): Unauthorized saving/learning copies.
• Right of public transmission / making available (Arts. 23, 23-2): Distribution of outputs (summaries, answers) and caches.
• TDM exception (Art. 47-7): Allows reproduction for machine learning under conditions. But output provision, secondary use, or scraping against site terms may fall outside its scope.
• Quotation exception: Requires subordinate use, clear source, necessity. If summaries/answers overshadow originals, it weakens legality.

Court will likely scrutinize: collection method (crawl & robots.txt compliance), storage form (duration, encryption, reuse), and output fidelity (degree of verbatim reproduction).
Note: Perplexity is also facing parallel lawsuits in U.S. courts (e.g., Dow Jones in NY), creating a cross-jurisdictional battle.

Perplexity’s Expected Defenses

• Invoke TDM exception: Learning copies are rights-limited; may contest Japanese law’s territorial application since processing occurs abroad.
• Outputs as transformation: Summaries/answers = transformation, not reproduction (qualitative/quantitative assessment).
• Licensing efforts: Stress new licensing/opt-out schemes to reduce perceived willful infringement (though this doesn’t cure past conduct).

Keys to Outcome

1. Scope of TDM exception: Learning copies may be legal, but output provision/secondary use may not be covered.
2. Similarity of outputs: Degree of verbatim overlap with original text.
3. Compliance with site rules: Ignoring robots.txt/ToS could factor as tort or contract breach.
4. Damage proof: Models to quantify lost readership/advertising revenue will matter.

Industry & Monetization Impact

• In Japan, expect three-tier licensing needs: for training, summarization, and distribution.
• Attribution and backlinking may become de facto standards.
• Legal logging (sources, timestamps, reproduction ratio) will likely be built into procurement/contract terms.

3) Comparison of the Two Lawsuits

4) Business & Investment Checklist (Practical Takeaways)

1. Distribution strategy: Estimate CAC/exposure under both native OS integration and external app models.
2. Licensing design: For sensitive data (e.g., news), prepare separate contracts/log trails for training, summarization, and distribution.
3. Product design: Automate attribution/backlinks/quotation compliance, enforce similarity thresholds.
4. Legal infrastructure: Enable auditable logs across model, data, distribution (collection path, TDM assessment, cache mgmt).
5. Risk diversification: Avoid overreliance on one OS or publisher; secure multiple entry points (Web/other OS/third parties) and multiple licensing sources.

References (Primary / Major Media)

• xAI vs Apple/OpenAI
  • Reuters (claims & relief): https://www.reuters.com/legal/litigation/elon-musks-xai-sues-apple-openai-over-ai-competition-app-store-rankings-2025-08-25/
  • Bloomberg (market definition & overview): https://www.bloomberg.com/news/articles/2025-08-25/musk-sues-apple-openai-over-alleged-ai-arrangement-on-iphones
  • CBS/AP (antitrust framing): https://www.cbsnews.com/news/elon-musk-apple-openai-antitrust-lawsuit/
• Nikkei/Asahi vs Perplexity
  • The Japan Times (court, damages, injunction): https://www.japantimes.co.jp/news/2025/08/26/japan/crime-legal/japan-newspapers-sue-ai-startup/
  • Financial Times (unauthorized copy/save claim): https://www.ft.com/content/79a88d1a-d914-4188-8792-0a20973b39a1
  • Related: Dow Jones suit in U.S. (motion to dismiss/transfer denied): https://www.reuters.com/legal/litigation/perplexity-ai-loses-bid-dismiss-or-transfer-news-corp-copyright-case-2025-08-21/

Note: This summary is based on currently available public reports. Evaluations may shift with amended complaints, responses, or injunction rulings. For decisions, consult primary documents (complaints, court orders) before making critical judgments.