In 2026, Brazilian newspaper Folha de S.Paulo signed a commercial agreement with Google allowing its Portuguese-language real-time news to be used in Google’s AI products, including Gemini. This followed a similar settlement Folha reached with OpenAI, demonstrating publishers’ strategy of playing both sides—suing some AI companies while licensing to others. The deal is part of Google’s broader global strategy to license publisher content for its AI systems, reducing legal risk and positioning Gemini as the “compliant” AI for enterprise customers concerned about copyright exposure.
Google’s approach contrasts sharply with OpenAI and Perplexity, which face dozens of copyright lawsuits alleging unauthorized training on protected content. By signing licenses upfront, Google accepts higher content costs but gains legal certainty, publisher goodwill, and a competitive differentiator for risk-sensitive enterprise buyers. The strategy also provides Gemini with access to fresh, licensed news content—improving answer quality for current events compared to models trained on static, potentially outdated datasets.
The Folha deal is emblematic of a broader pattern. Google has signed similar agreements in multiple countries (US, UK, Germany, Australia, Brazil) with major publishers including Associated Press, Axel Springer, News Corp (after initial litigation), and regional newspapers. This coordinated global strategy secures content rights before regulators or courts force the issue, creating a competitive moat: Google can claim its AI is trained on properly licensed content while competitors face injunctions or damages that constrain their products.
Deal Announcement: 2026 (specific dates vary by publisher)
Key Partner: Folha de S.Paulo (Brazil’s largest newspaper, Portuguese-language content)
Content Scope: Real-time news articles for Gemini AI products
Geographic Scope: Global licensing program spanning US, Europe, Latin America, Asia-Pacific
Strategic Contrast: Google licenses upfront vs. OpenAI/Perplexity “deploy first, litigate later”
Enterprise Benefit: Reduced copyright liability for Gemini enterprise customers
Google Publisher Licensing Strategy
+--------------------------------------------------------------------------+
| Google Gemini Publisher Licensing Flywheel |
+--------------------------------------------------------------------------+
[Google Negotiates with Publishers]
- Offer licensing fees for AI training rights
- Promise attribution and traffic referrals
│
┌────────┴──────────────────────────────────────────┐
▼ ▼
[Publishers Agree to License] [Google Integrates Content]
- Legal certainty, guaranteed revenue - Fresh news in Gemini answers
- Avoid costly litigation - Improved answer quality
- Traffic from AI referrals - Competitive differentiation
│ │
└──────────────────┬────────────────────────┘
▼
[Google Benefits]
- Reduced legal risk vs. competitors
- Enterprise sales advantage
- Better AI answer quality
- RESULT: Sustainable content moat
+--------------------------------------------------------------------------+
AI Company Content Strategy Comparison
The table below compares how major AI companies approach publisher content:
Company
Strategy
Key Deals
Legal Exposure
Enterprise Positioning
Google (Gemini)
Proactive licensing upfront
AP, Axel Springer, Folha, News Corp (post-litigation)
Low (licensed content)
“Compliant AI for enterprise”
OpenAI (ChatGPT)
Deploy first, negotiate/litigate later
Some deals (unknown terms), but 30+ lawsuits pending
High (unlicensed training)
“Best capabilities, accept legal risk”
Perplexity
Aggressive scraping, fight fair use
Few licenses, many lawsuits (CNN, News Corp, Folha)
Very high (core product challenged)
“AI search despite publisher opposition”
Anthropic (Claude)
Mixed: some licensing, cautious deployment
Limited public deals, fewer lawsuits than OpenAI
Medium (more conservative training)
“Safety-first, lower legal risk”
Microsoft (Copilot)
Leverage existing licenses + new deals
News Corp, AP, some via OpenAI partnership
Medium (depends on data source)
“Enterprise-grade with Microsoft indemnification”
Google (Gemini)
Strategy Proactive licensing upfront
Key Deals AP, Axel Springer, Folha, News Corp (post-litigation)
Legal Exposure Low (licensed content)
Enterprise Positioning “Compliant AI for enterprise”
OpenAI (ChatGPT)
Strategy Deploy first, negotiate/litigate later
Key Deals Some deals (unknown terms), but 30+ lawsuits pending
Legal Exposure High (unlicensed training)
Enterprise Positioning “Best capabilities, accept legal risk”
Perplexity
Strategy Aggressive scraping, fight fair use
Key Deals Few licenses, many lawsuits (CNN, News Corp, Folha)
Legal Exposure Very high (core product challenged)
Enterprise Positioning “AI search despite publisher opposition”
Anthropic (Claude)
Strategy Mixed: some licensing, cautious deployment
Key Deals Limited public deals, fewer lawsuits than OpenAI
Legal Exposure Medium (more conservative training)
Enterprise Positioning “Safety-first, lower legal risk”
Microsoft (Copilot)
Strategy Leverage existing licenses + new deals
Key Deals News Corp, AP, some via OpenAI partnership
Legal Exposure Medium (depends on data source)
Enterprise Positioning “Enterprise-grade with Microsoft indemnification”
Real-World Utility & Policy Implementation
The 4-Step AI Content Licensing Playbook
Inventory Content Categories by Risk: Classify training data sources by copyright risk: high-risk (news publishers, books, stock imagery, code), medium-risk (blogs, forums, social media), low-risk (public domain, Creative Commons, openly licensed). Prioritize licensing negotiations for high-risk categories where lawsuits are most likely and damages highest.
Develop Licensing Framework Templates: Create standardized licensing agreements for different content types (news, books, images, code) with flexible terms: flat annual fees, per-query royalties, revenue-sharing, or hybrid models. Include attribution requirements, usage reporting, and audit rights. Having templates ready accelerates negotiations and demonstrates seriousness to publishers.
Build Attribution and Analytics Infrastructure: Implement technical systems to track which licensed content is used in AI outputs, enabling accurate royalty calculations and publisher reporting. Provide publishers with dashboards showing how their content is used in AI answers, traffic referrals generated, and revenue earned. Transparency builds trust and reduces publisher incentives to litigate.
Communicate Licensing to Enterprise Customers: For enterprise sales, highlight licensed content as a differentiator: “Gemini is trained on properly licensed news, reducing your copyright liability vs. competitors.” Include indemnification clauses in enterprise contracts that protect customers from publisher lawsuits. This positions licensing as a risk mitigation feature, not just a cost center.
Licensing Creates Competitive Moat: Google’s proactive licensing strategy reduces legal risk and positions Gemini as the “safe” choice for enterprises concerned about copyright exposure. As publisher lawsuits against OpenAI and Perplexity escalate, expect enterprise buyers to factor legal risk into AI vendor selection—potentially choosing Gemini despite slightly lower capabilities or higher costs.
Content Quality Improves with Fresh Data: Licensed real-time news access gives Gemini an advantage in answering questions about current events. Models trained on static datasets (with knowledge cutoffs) cannot reliably answer questions about recent news without risking hallucination. Licensed news feeds enable Gemini to provide accurate, sourced answers about breaking news—a key differentiator for research and analysis use cases.
Publishers Play Both Sides Strategically: Folha’s deals with both Google and OpenAI (after settling) show publishers will license to some AI companies while suing others. This creates a fragmented landscape where AI companies must negotiate individually with each major publisher—no industry-wide licensing framework exists yet. Expect this to continue until courts establish clearer fair use precedent, at which point licensing terms may standardize.
Everyday Applications & Decision Framework for Readers
The commercial licensing pact between Google and Folha de S.Paulo offers clear lessons for everyday digital professionals, media consumers, and business operators navigating the evolving artificial intelligence landscape.
How Everyday Consumers Can Verify Licensed AI Citations
When using Gemini, Google Search AI Overviews, or ChatGPT, everyday users must recognize whether an answer stems from static training data, unverified web scrapes, or licensed real-time journalism. In practice, you can apply three verification steps in your daily research:
Examine Source Grounding Links: In Gemini, click the source pill icons at the bottom of the response. Answers grounded in direct publisher agreements feature direct links to canonical newsrooms rather than aggregator domains.
Cross-Check Breaking Event Timestamps: Unlicensed models often hallucinate details on events occurring within the past 48 hours. When researching financial earnings, regulatory rulings, or international elections, verify that the AI cites news published within hours of the query.
Use Primary Confirmation for High-Stakes Decisions: If using AI summaries for investment, legal, or health decisions, treat the AI summary as an index and read the primary publisher source directly before taking action.
Practical Playbook for Small Business Owners & Content Creators
For independent publishers, technical bloggers, and digital agency leaders, the rise of multi-million-dollar publisher licensing deals creates immediate operational decisions. You do not need a billion-dollar catalog to protect and monetize your intellectual property:
Implement Granular Bot Controls: Audit your server configuration to differentiate between search indexers (Googlebot) and training crawlers (Google-Extended, GPTBot, ClaudeBot). If your business relies on ad revenue or subscriptions, restricting training bots while preserving search bots protects your content value.
Publish Proprietary Formats: AI scrapers excel at summarizing commodity text. To maintain reader retention, integrate downloadable worksheets, interactive calculators, podcast interviews, and firsthand case studies that cannot be replicated in a single chat prompt.
Leverage Enterprise Indemnification: If your company builds customer-facing tools using the Gemini API via Google Cloud Vertex AI, verify that your enterprise contract includes Google’s copyright indemnity clause, shielding your organization from third-party copyright claims.