What is the role of news publishers in AI?

The Critical Role of News Publishers in AI Development and Deployment

News publishers serve as essential content providers and stakeholders in the artificial intelligence ecosystem, shaping how AI models are trained, deployed, and regulated. Their role extends far beyond simply providing raw data—publishers actively negotiate licensing agreements, advocate for fair compensation, and work to establish industry standards for attribution and citation in AI-generated content. Understanding this multifaceted role is crucial for anyone interested in how AI systems access, process, and present journalistic content to users worldwide.

Providing High-Quality Training Data for AI Models

News publishers supply the foundational training data that powers modern AI language models and search systems. Major news organizations produce vast amounts of professionally edited, fact-checked, and well-structured content that AI developers find invaluable for training purposes. This content includes news articles, investigative reports, opinion pieces, and multimedia materials that help AI models understand language patterns, current events, and complex topics with greater accuracy and nuance than unvetted internet content alone.

The quality of journalistic content makes it particularly valuable for AI training. News publishers employ editorial teams, fact-checkers, and subject matter experts who ensure accuracy and reliability—qualities that directly improve AI model performance. When AI companies train their models on news content, they benefit from decades of journalistic standards and professional writing practices. This relationship has become so important that major AI companies like Amazon, Meta, and OpenAI have actively pursued licensing agreements with leading publishers including The New York Times, News Corp, and USA Today to secure access to their content libraries.

Negotiating Content Licensing and Compensation Agreements

The landscape of publisher-AI company relationships has evolved significantly, with publishers now negotiating sophisticated licensing agreements that define how their content can be used. Initially, when generative AI systems first emerged in late 2022, publishers faced a challenging situation where their content had already been incorporated into AI models without explicit permission or compensation. This prompted a wave of licensing negotiations that fundamentally changed how AI companies and publishers interact.

Early licensing deals typically involved one-time lump sum payments for training data access. For example, Amazon agreed to pay The New York Times between $20 million and $25 million annually under a multi-year content licensing agreement, while News Corp secured approximately $50 million for similar arrangements. However, the industry has rapidly evolved beyond these initial training-focused deals. Publishers and AI companies have increasingly shifted toward usage-based licensing models, particularly those centered on “AI grounding” or Retrieval Augmented Generation (RAG) technology.

Advocating for Proper Attribution and Citation Standards

News publishers have become vocal advocates for accurate attribution and citation practices in AI-generated content, recognizing that proper credit directly impacts their traffic, brand visibility, and revenue generation. Research from the Tow Center for Digital Journalism revealed that over 60% of AI-generated responses contain incorrect or misleading information, and many AI search tools fail to properly attribute sources or cite original publishers.

A critical issue publishers face is that AI search engines often cite syndicated or republished versions of articles rather than crediting the original news organization that broke the story. This practice diminishes visibility for primary publishers and deprives them of direct referral traffic. Some AI platforms, including Grok and Gemini, have been documented generating broken or fabricated URLs, further reducing traffic to legitimate news sites. Publishers argue that proper attribution should include direct links back to their original articles, not secondary sources or aggregators.

The News Media Alliance has developed an AI Licensing Program specifically to address these concerns, promoting efficient marketplace solutions that ensure publishers receive appropriate credit and compensation. Industry groups continue to advocate for stronger AI regulations that would require transparent policies mandating proper citation and linking practices. These efforts represent publishers’ attempt to establish industry-wide standards that protect journalistic integrity while enabling AI systems to function effectively.

Shaping AI Search Engine Behavior and Visibility

Publishers influence how AI search engines operate through their licensing agreements and content control mechanisms. When publishers negotiate with AI companies, they can establish terms that affect how their content appears in AI-generated answers, whether it receives proper attribution, and how frequently it can be accessed. These negotiations directly shape the user experience in AI search tools like Perplexity, Google AI Overviews, ChatGPT, and Claude.

However, publishers face ongoing challenges in enforcing content controls. Many AI platforms routinely retrieve content from publisher websites even when publishers explicitly block them using robots.txt, a standard technical tool for controlling web crawling. This disregard for publisher restrictions raises ethical concerns and undermines publishers’ ability to manage how their content is used. Some publishers with formal partnerships with AI companies still experience misattribution or see their content surface in ways that don’t drive traffic back to their platforms, suggesting that agreements alone are insufficient without proper enforcement mechanisms.

Addressing Copyright and Intellectual Property Concerns

News publishers have raised significant copyright and intellectual property questions regarding AI training on their content without explicit permission or compensation. The U.S. Copyright Office has examined whether copyrighted material can be used to train AI systems, recognizing that copyright law protects intellectual creations including newspaper articles, subject to certain exceptions. Publishers argue that their original reporting represents valuable intellectual property that should not be freely exploited by AI companies.

These copyright concerns have prompted legal action and regulatory scrutiny. Publishers contend that AI companies have essentially “stripped” their content to train models without adequate compensation or permission. This has led to ongoing litigation and policy discussions about fair use, licensing requirements, and the appropriate compensation models for AI training. The resolution of these copyright questions will significantly impact how publishers and AI companies interact in the future and whether publishers can effectively control and monetize their content in AI systems.

Influencing AI Regulation and Industry Standards

News publishers actively participate in shaping AI regulation and industry standards through industry groups, policy advocacy, and direct engagement with regulators. Organizations like the News Media Alliance, Digital Content Next, and individual publishers work with policymakers to develop frameworks that protect journalistic interests while enabling responsible AI development. Publishers advocate for regulations that would require AI companies to obtain explicit permission before using copyrighted content, provide transparent attribution, and establish fair compensation mechanisms.

Publishers also influence emerging industry standards through their participation in technical working groups and standards bodies. The IAB Tech Lab, for example, is developing standardized frameworks for pay-per-crawl and pay-per-query models with input from publishers and AI companies. These collaborative efforts aim to create consistent, fair practices across the industry rather than relying on individual negotiations. As AI technology continues to evolve, publishers’ voices in these discussions become increasingly important for ensuring that journalistic content is treated fairly and that quality journalism remains economically viable.

Managing the Impact on Traffic, Engagement, and Revenue

News publishers must navigate the complex challenge of AI search disrupting their traditional traffic and revenue models while simultaneously leveraging AI as a distribution channel. Traditional search engines drive referral traffic to news sites, supporting subscription models, advertising revenue, and brand visibility. However, AI search tools that provide comprehensive answers without requiring users to visit source websites reduce the need for readers to click through to full articles, limiting publishers’ direct audience engagement opportunities.

This shift in user behavior directly threatens publisher revenue streams. When AI systems summarize news content without proper attribution or links, readers may never visit the publisher’s website, eliminating opportunities for subscription conversions, ad impressions, and brand engagement. Publishers report that AI-driven search changes user behavior by reducing the incentive to visit source websites, fundamentally challenging established business models. To address this challenge, publishers are developing AI-optimized content strategies, similar to how they adapted to search engine optimization (SEO) decades earlier, exploring ways to maximize visibility and ensure their content drives traffic in an AI-driven search environment.

Collaborating on Content Partnerships and Distribution

Forward-thinking publishers are moving beyond adversarial relationships with AI companies to establish collaborative partnerships that create mutual value. Rather than simply licensing historical content for training, publishers are increasingly partnering with AI platforms to ensure their latest reporting reaches AI users in real-time. These partnerships often include revenue-sharing arrangements where publishers benefit when their content is cited in AI-generated answers.

Perplexity’s Publisher Program exemplifies this collaborative approach, incorporating Retrieval Augmented Generation (RAG) technology to include trusted publisher content in answers while providing attribution and revenue sharing. Gannett’s partnership with Perplexity, which includes USA Today and the USA Today Network, demonstrates how publishers can negotiate terms that ensure their content receives proper visibility and drives value. These collaborative models suggest a future where publishers and AI companies work together to create better user experiences while ensuring publishers receive appropriate compensation and attribution for their content.