Google CEO Sundar Pichai outlines content creators’ vision for compensation in December at the New York Times’ Dealbook Summit, suggesting the emergence of a marketplace where creators develop content exclusively for artificial intelligence platforms. His comments came as publishers report significant traffic losses from Google’s AI-powered search capabilities.
According to an article published by the Atlantic, Pichai showed that market forces determine compensation structures without necessarily including traditional publishers. “I think there will be a market in the future. I think there will be creators who will create for AI,” Pichai said at the summit. “People will understand that.”
Get the PPC Land Newsletter For more details like this.
Subscribe
summary
WHO: major news publishers such as Google CEO Sundar Pichai and Dotdash Meredith, Daily Mail, The Atlantic, and The News Media Alliance
Content: Pichai proposed at the New York Times dealbook summit that a market will emerge for creators to develop content for AI platforms, but publishers report significant traffic losses from Google’s AI-powered search capabilities
Time: Pichai’s comments occurred during the dealbook summit in December 2025. This occurred amid ongoing tensions throughout 2025 as Google expanded its AI mode access and publishers recorded a decline in traffic.
Location: The concept of marketplace addresses global content creation, with current competition focusing primarily on the US and European markets where Google’s AI features have been launched
Why: The situation is important to the marketing community as AI-powered search capabilities can essentially allow users to discover content and create new challenges for content creators who rely on revenue search traffic while transforming digital marketing strategies and website monetization models.
Publishers face pressure from AI capabilities
The statement comes amid growing tensions between Google and major news organizations over AI-powered search capabilities. According to Ahrefs’ comprehensive analysis, Google’s AI overview, when displayed in search results, reduces traffic to external websites by more than 34%. This study compared the same period in 2025 with 300,000 searches in March 2024.
Dotdash Meredith, publishing people, better homes and gardens, food and wine, has begun preparing what CEO Neil Vogel calls the “Google Zero” scenario. According to the Atlantic Report, some industry observers speculate that traffic from chatbots has slowed down, contributing to recent layoffs at outlets such as Business Insider and The Daily Dot.
Rich Caccappolo, vice-chairman of media for the company that publishes the Daily Mail, described the threat in harsh words during an interview with the Atlantic. “It’s absolutely terrifying,” he said when asked if the AI-generated answer could eliminate his company. “And my concern is that it won’t happen in three or five years. I’m kidding that it’ll happen next Tuesday.”
Training data creates unbalanced negotiations
The remuneration debate comes from what publishers describe as systematically unfair licensing negotiations. According to Atlantic Report, at least 72 license transactions have been completed between publishers and AI companies over the past two years. However, Capcappolo told the Atlantic that he “feels an incredible imbalance at the negotiation table.”
This article reveals that each book can only be licensed a few hundred dollars. AI companies may then use the materials without payment, and publishers requesting higher compensation may be denied. This dynamic stems from what the report describes as demonstrating AI companies’ “ability and willingness to take it without paying.”
Technical developments increase concerns
As of May 1, 2025, Google has extended AI mode access to all US users and eliminated the previous waitlist system that restricts selected tests. This feature provides a conversation search experience through the “query fan-out technique” that simultaneously generates multiple related searches across sources, and what Google describes.
According to Google Docs, AI Mode integrates shopping graph data with over 45 billion product lists updated over 2 billion times per hour. The system integrates information from multiple sources to create a comprehensive response, reducing user incentives and access individual websites for more information.
The News Media Alliance views Google’s approach as a “definition of theft,” according to a statement from President and CEO Danielle Coffey on May 21. The organization calls for the Ministry of Justice’s intervention to represent an escalation of a direct legal threat.
Marketplace Vision is in contrast to current practices
Pichai’s Marketplace concept is very different from current AI training practices. According to an Atlantic report, Openai CEO Sam Altman previously proposed an “opt-in” regime that allows authors to receive “micropayments” when work affects AI output. However, this approach contrasts with Openai’s current practice of using artist and writer styles without compensation or effective opt-out mechanisms.
The market model Pichai coincides with Silicon Valley’s historical approach to industry transformation. According to Atlantic analysis, tech companies usually view established institutions as intermediaries that are excluded for efficient profits. This pattern emerged on ride-sharing platforms that first attracted drivers before they could control pay, benefits and workloads algorithmically.
Legal and technical barriers complicate solutions
Publishers face major challenges in preventing content misuse. According to Atlantic Report, the robot exclusion protocol (a standard opt-out method available to news publishers) is easily avoided. AI companies generally maintain secrets about their training data, making it difficult for publishers to identify which companies are using their content.
Law scholar James Grimmelman told the Atlantic that AI companies can protect licensing transactions by covering more than training rights. Companies may argue that they will indemnify clean article versions, real-time content feeds, or liability releases for chatbot plagiarism rather than access to training materials.
At least 12 lawsuits have been filed against AI companies, including more than 20 publishers. However, according to Atlantic Report, these cases can only be resolved after irreparable damage has occurred to the disclosure of the business model.
The exchange of revenue remains unknown
The license agreement provides minimal compensation compared to lost advertising and subscription revenue. According to sources from the publisher of Atlantic Report, we cannot approach lost revenue from the decline in readers caused by AI-powered search capabilities.
Some publishers have pursued a dual strategy by combining legal action with licensing negotiations. Atlantic itself maintains and cooperates with Open Eye at the same time. This approach reflects uncertainty about which strategies are most effective in protecting publishers’ interests.
Digital media company Ziff Davis analyzed a web-based AI training dataset and determined that content from “high authority” sources is more valuable to AI companies than blog posts and social media content. Microsoft researchers have publicly emphasized the “importance of high-quality data,” and textbook-style content may be particularly desirable for AI training.
Industry transformation accelerates
Google spent $75 billion on AI infrastructure in 2025. This represents a significant increase from $2 billion a few years ago, according to the arrival of Pichai’s Bloomberg Tech Summit on June 5th.
According to a recent statement to various media outlets from Pichai, AI eliminates friction between content formats through native multimodal capabilities. “AI will zero friction to move from one format to another,” he explained in an interview on May 27, 2025, following Google’s I/O developer conference.
The CEO highlighted existing features including an audio overview of notebook elements that can generate podcast-style content and process multiple documents. This technology could allow for automatic conversion between text, audio, video and interactive formats without the additional production costs of content creators.
Alternate platforms appear
Advertising technology company Taboola launched Deeperdive on June 11, 2025, positioning its platform as “the industry’s first Gen AI Answer Engine designed to live directly on the publisher’s website.” Major Media Organizations Gannet | USA Today Network and The Independent joined as Launch Partners.
DeeperDive works by embedding AI-powered search capabilities in publisher websites rather than redirecting traffic elsewhere. According to Adam Singolda, CEO of Taboola, “Open Web thrives when innovation and equity are closely linked. It is not sustainable for the GENAI engine to send most of the traffic and compensate only for the selected few while relying on publisher content.”
The platform leverages data from Taboola’s network of 600 million active users every day, spanning 9,000 publisher partners. This approach contrasts with traditional AI search engines that aggregate content from multiple sources while keeping users within their own interface.
Timeline
